3 : ^> I North C: 3 Library
DEPARTMENT OF CONSERVATION AND DEVELOPMENT
GEORGE R. ROSS, DIRECTOR
DIVISION OF MINERAL RESOURCES
JASPER L. STUCKEY, STATE GEOLOGIST
Bulletin Number 61
in North Carolina
LEWIS J. HASH
EARL C. YAN HORN
KEFTON H. TEAGUE
PREPARED IN COOPERATION WITH THE TENNESSEE VALLEY AUTHORITY
Department of Conservation and Development
George R. Ross, Director
Division of Mineral Resources
Jasper L. Stuckey, State Geologist
Bulletin Number 61
in North Carolina
Lewis J. Hash
Earl C. Van Horn
Kefton H. Teague
North Carolina State Library
Raleigh, N. Q
Prepared in Cooperation with the Tennessee Valley Authority
MEMBERS OF THE BOARD OF CONSERVATION
Governor W. Kerr Scott, Honorary Chairman Raleigh
Miles J. Smith, Chairman Salisbury
Walter J. Damtoft, Vice Chairman Canton
Chas. S. Allen Durham
W. B. Austin Jefferson
Oscar P. Breece Fayetteville
Aubrey L. Cavenaugh Warsaw
Staley A. Cook Burlington
Ferd Davis Zebulon
C. Sylvester Green Chapel Hill
Fred P. Latham Belhaven
Mrs. Roland McClamroch Chapel Hill
J. C. Murdock Troutmans
W. Locke Robinson Mars Hill
Eric W. Rodgers . Scotland Neck
George R. Ross, Director
LETTER OF TRANSMITTAL
Raleigh, North Carolina
April 19, 1951
To His Excellency, Honorable W. Kerr Scott
Governor of North Carolina
I have the honor to submit herewith manuscript for publica-
tion as Bulletin 61, "Sillimanite Deposits of North Carolina."
This Bulletin is another in a series being made possible by the
cooperation of the Tennessee Valley Authority.
Considerable interest has been developed in sillimanite and
other aluminum silicate minerals in recent years. It is believed
that this report will be of considerable value to those interested
in aluminum silicate minerals.
George R. Ross,
Purpose and Scope 1
Field Work and Acknowledgments 1
PROPERTIES OF SILLIMANITE 2
USES OF SILLIMANITE 2
SILLIMANITE IN THE CLIFFSIDE-ELKIN BELT 3
Physical Geography 3
Description of Deposits 6
Polkville Deposit 6
Wards Creek Deposit 6
Casar Deposit 6
Smith Cliff Deposit 8
Saw Mills Deposit 10
Cages Mountain Deposit 10
Dudley Shoals No. I Deposit : 12
Dudley Shoals No. II Deposit „ 14
Dudley Shoals No. Ill Deposit 16
Ellendale School Area 16
Fox's Orchard Deposit 16
Wilkesboro-Taylorsville Area 17
SILLIMANITE IN THE WARNE-SYLVA BELT 17
Physical Geography 17
Descriptions of Deposits 20
Brasstown Church Cherry Mountain Deposit 20
Hyatt Mill Creek Deposits 23
Downing Creek Deposits 23
Sap Sucker-Cold Branch Deposits 25
Tusquitee Gap-Jarrett Creek Deposits 25
Burningtown Deposits 26
Saldeer Deposits 26
Oak Grove Deposits 28
CONTENTS — CONTINUED
Etna-Bradley Creek Deposits 28
Rickman Creek Deposits - 28
Matlock Creek Deposit ... 30
Leatherman Deposits 30
Cowee Church Deposits 30
Greens Creek School Deposits 30
Brook Branch-Sutton Branch Deposits 32
Riverview Church Deposits 32
Sylva Deposits 32
OTHER SILLIMANITE DEPOSITS 35
Grassy-Ridge-Big Ridge Deposits 35
South Hominy Deposit 35
Asheville-Mount Mitchell Deposits 36
LABORATORY RESULTS 36
Method Employed for Sample Analysis 36
Analysis and Location of Samples from the Cliffside-
Elkin Sillimanite Belt 37
Description of Heavy Minerals in Sillimanite Ore
From Cliffside-Elkin Belt 38
Analysis and Location of Samples From the Warne-Sylva
Sillimanite Belt I 41
Description of Heavy Minerals in Sillimanite Ore
From Warne-Sylva Belt _ 42
Introduction , 45
Cages Mountain Deposit 45
Fox's Orchard Deposit 46
Smith-Cliff Deposit 46
Dudley Shoals Deposits 47
Grassy Ridge Deposit . 48
PROSPECTING, MINING AND RESERVES 49
POSSIBILITIES FOR DEVELOPMENT 50
1. Occurrence of Sillimanite in Cliffside-Elkin Belt
of North Carolina l ^ ^ -.-. 7
2. Smith Cliff Sillimanite Deposit, Burke County, North Carolina 9
3. Valdese-Dudley Shoals Sillimanite Area, Burke, Caldwell,
Alexander Counties, North Carolina 11
4. Cages Mountain Sillimanite Deposit, Caldwell County,
North Carolina .__ .. 13
5. Dudley Shoals No. I Sillimanite Deposit, Caldwell
County, North Carolina 15
6. Occurrences of Sillimanite in the Warne-Sylva Belt of
North Carolina ill. 1 ..... 19
7. Hayesville Area: Brasstown Church-Cherry Mountain
Sillimanite Deposits 21
8. Hayesville Area: Hyatt Mill Creek Sillimanite Deposits 22
9. Hayesville Area : Downing Creek Sillimanite Deposits , 24
10. Etna-Cowee Area : Saldeer Sillimanite Deposits . 27
11. Etna-Cowee Area : Oak Grove-Matlock Creek Sillimanite
Deposits L . 29
12. Etna-Cowee Area : Leatherman Sillimanite Deposits 31
13. Greens Creek Area Sillimanite Deposits . 33
14. Sylva Area Sillimanite Deposits .^ _i ciiil : 34
PURPOSE AND SCOPE
This report describes the principal sillimanite deposits in North Carolina. In the first comprehensive
report on a new mineral it is desirable to give a complete description of the geology of the mineral, but
time allotted to the study did not permit detailed geological studies ; therefore, this phase of the report has
been kept to a minimum. An attempt has been made to assemble the data in such form that they would be
of use to those who are interested in prospecting, mining, and using sillimanite.
The reconnaissance field work has shown that sillimanite occurs in most all counties west of the cen-
tral Piedmont area of the State. The major sillimanite deposits and those offering the best, production
possibilities occur in two belts, one located in the upper Piedmont, designated "the Cliffside-Elkin belt",
and the other located west of the Blue Ridge designated "the Warne-Sylva belt." The authors are aware
of sillimanite occurrences outside of these two principal belts, but most of these other known occurrences
are small and appear to be of little economic importance. The two principal belts are so extensive that it
was not feasible to obtain all the desired data.
Since the deposits are spread over large areas, one of the problems confronting those interested in the
sillimanite has been that of determining which deposits are best suited for mining and concentration. It
is hoped that the data presented in this report will serve to delimit the deposits so that the more promising
ones can be more readily selected.
Laboratory data, regarding the percentage of sillimanite in and methods of beneficiation of the vari-
ous ores, have been included in this report to show the possible methods of treatment and type of concen-
trates that can be produced from North Carolina deposits.
Prior to 1900 sillimanite grains were identified in concentrates from placer operations for monazite 7 *
(1915) and corundum 8 (1905), however, this early published literature made no mention of the primary
source of the sillimanite. During the early part of 1944 Charles E. Hunter of the Tennessee Valley Author-
ity noted extensive sillimanite schist occurrences in the vicinity of Valdese, Burke County, North Carolina.
In the summer of 1945 Hunter and White 3 made a reconnaissance survey and found that the sillimanite
deposits in the upper Piedmont area extended from the South Carolina line northeastwardly, to the Yadkin
River in the vicinity of North Wilkesboro and Elkin.
Probably the first reference to sillimanite in the "Warne-Sylva belt" was that of Prindle 9 who men-
tioned occurrences in the vicinity of Hayesville, Clay County. In 1945 Furcron and Teague 2 described sil-
limanite in the same vicinity, referring to unpublished field studies of J. L. Calver and W. T. McDaniel, Jr.,
Geologists, TVA, who made reconnaissance studies of Clay County sillimanite deposits in 1944.
To date, there has been no production of sillimanite from the North Carolina deposits, however, since
the discovery there has been continued interest in these occurrences by various individuals and mining com-
FIELD WORK AND ACKNOWLEDGMENTS
The field work for this report was conducted as a joint project between the North Carolina Division of
Mineral Resources and the Regional Minerals Section of the Tennessee Valley Authority. General super-
vision of the studies was under Jasper L. Stuckey, State Geologist of North Carolina, and the late H. S.
Rankin, Senior Mining Engineer of TVA. The work was under the direct supervision of Charles E. Hunt-
er and details of the survey were conducted by Earl C. Van Horn and Lewis J. Hash, Geologists of the
TVA. Field assistants provided by the North Carolina Division of Mineral Resources included P. A.
Hager, J. H. Recknagel, W. C. Cole, and J. H. Davis.
The percentage of sillimanite and petrographic characteristics of the field samples were determined by
Euan K. Green, Student Assistant of the North Carolina Division of Mineral Resources. The field survey
♦References at end of report.
2 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
was started in June 1949 and completed in September of the same year. The data regarding beneficiation
of the sillimanite ore were obtained from laboratory studies by W. T. McDaniel, Jr. and Mason K. Banks,
Mining Engineers of TVA, and W. H. Powell, Jr., Ore Dressing Engineer of the North Carolina State Col-
lege Minerals Research Laboratory. Chemical analyses were prepared by Phillip N. Sales, Chemist, North
Carolina State College Minerals Research Laboratory. Maps included in this report were prepared by the
Maps and Surveys Branch of TVA.
PROPERTIES OF SILLIMANITE
Sillimanite, like kyanite and andalusite, has the theoretical chemical composition of Al 2 3 .Si0 2 . It has
a vitreous-silky to a subadamantine luster, specific gravity of 3.2-3.3 and is gray to bluish gray in color.
The hardness varies between 6 and 7 in Mohs' scale. Under the binocular microscope the crystals are trans-
parent. The mineral may occur as needles (fibrolite), composed of fibrous sometimes radiating hair-like
crystals, in schist or as coarse bundles of crystals (nodules) disseminated in schist.
When heated above 1650° C. sillimanite expands about 6.5 percent and is converted into a stable mix-
ture consisting of mullite (3Al 2 3 .Si0 2 ) and vitreous silica, with a specific gravity of about 3.15.
In the Warne-Sylva belt part of the sillimanite apparently has been formed from the alteration of kyanite
to sillimanite, however, the identity of these two minerals can be separated readily in the field by a simple
test of hardness. Also most sillimanite shows sericitization, however, this type of alteration can too be
readily detected by a hardness test.
USES OF SILLIMANITE
Since sillimanite has never been produced on a commercial scale there are no established uses for the
mineral ; however, as its various properties including physical, chemical, and thermal are known, certain
potential uses can be listed.
Sillimanite shows no apparent change when heated below 1600° C, thus it can be incorporated into
refractory bodies for service below that temperature. For service in excess of 1650° C. sillimanite is con-
verted into mullite. Some of the potential uses for sillimanite include porcelain for spark plugs, high-
alumina refractory brick, crucibles, saggers of all types, high temperature cements, linings for indirect-arc
and heat-treating furnaces, pyrometer tubes, and glass-tank blocks.
Research directed toward the use of sillimanite in porcelain has been conducted at Clemson College,
South Carolina. 11 , 12 Results from these investigations indicate that when sillimanite is substituted for
flint in porcelain bodies, strength is increased greatly, resistance to shock, etc. is improved, thus opening a
field for finely-ground sillimanite in the ceramic porcelain industries.
Considerable difference of opinion exists as to the importance of regional and thermal metamorphism,
hydrothermal action, composition of igneous intrusions, composition of original sediments, and character of
assimilation in the formation of sillimanite. Insufficient data are available upon the sillimanite-bearing
rocks and adjacent surrounding rocks of the North Carolina sillimanite deposits to permit a positive state-
ment as to their origin.
Furcron and Teague 2 in discussing the sillimanite deposits in Hart, Elbert, and Madison Counties, Geor-
gia, state that "without doubt the original sediment has been altered by large scale folding and regional
metamorphism, but these rocks, representing the folded roots of the old sedimentary formations, are all alter-
ed and more coarsely crystallized than are the rocks to the west which have also suffered the same regional
effects. The recrystallization of the muscovite, biotite and sillimanite is believed to be due to the intimate
relation of the rock here to the intrusive granites, pegmatitic granites, and pegmatites." In studying de-
posits in the Davy Mountain area of Georgia which are the southwestern continuation of the "Warne-Sylva
belt" of this report, these same workers state that "the sillimanite occurs in zones of muscovite-graphite
schist which were probably highly aluminous, representing an old meta-sedimentary facies of the pre-Cam-
brian gneiss .... Some specimens from Davy Mountain and other localities still retain the structure and
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 3
outward appearance of kyanite, thus it appears that locally, at least, kyanite produced by hydrothermal al-
teration of the schist was converted to sillamanite through the action of later, hotter pegmatite solutions."
Smith 13 who has studied the sillimanite deposits in South Carolina states "Sillimanite has resulted from
contact metamorphism of a schist by granite intrusions .... There is no evidence that any new constitu-
ents have been added to the schist by igneous metamorphism. The sillimanite has apparently been devel-
oped by the agents of heat and aqueous solutions."
Sillimanite in North Carolina is not confined entirely to the highly argillaceous rocks (now schist),
but occurs also in quartzite. Chemical analyses of sillimanite rock from neither of the two belts are avail-
able. Criteria relating to the origin of the sillimanite are based, therefore, on physical distribution of the
sillimanite and its associated minerals. Structural control in the localization of sillimanite is quite evident
both in the Warne-Sylva belt and in the Cliffside-Elkin belt. In addition to structural features, the deposits
exhibit uniformity in that all are associated with more or less intense pegmatization.
As to mineral associations in the sillimanite deposits, muscovite, sericite, biotite, and quartz are present
without exception. Kyanite, garnet, staurolite, zircon, rutile, ilmenite, barite, and graphite are evident in
most of the sillimanite-bearing rock and may be present in all. Tourmaline, sphene, clinozoisite, and chro-
mite have been identified from some of the deposits.
Kyanite and garnet are sometimes abundant in both of the two principal sillimanite belts where silli-
manite is absent or present only in small amounts. Locally, sillimanite pseudomorphs kyanite and probably
replaces and is replaced by muscovite and quartz. These and other criteria are related to conceptions of
zone metamorphism and metamorphic facies which have been reviewed recently by such writers as Moore,
Ramberg, 10 Turner, 15 and Weiss. 16 In the North Carolina deposits available evidence does not indicate that
the sillimanite was formed in zones of great depth. The high degree of metamorphism is more closely re-
lated to intense igneous activity.
It is believed that the ancestral rocks, containing most of the materials necessary to form the present
mineral assemblages, were thoroughly foliated and sheared. Following shearing or during its last stages
pegmatites were introduced into the fractured and broken rock. The excess of alumina necessary for the
formation of sillimanite was derived presumably from the schist and concentrated by the pegmatites under
pressure. The rock evidently did not contain sufficient potash, soda, or lime to use all the available alumina
to form feldspar or muscovite.
After the formation of sillimanite later solutions were active in altering the sillimanite. Some of the
deposits contain much sericite which has pseudomorphed sillimanite.
SILLIMANITE IN THE CLIFFSIDE-ELKIN BELT
The geology and mineralogy of the two principal sillimanite belts are somewhat similar, but vary great-
ly in detail. Complete descriptions of each belt are contained in separate chapters, except for the descrip-
tions of laboratory tests. Direct supervision for the studies of the Cliffside-Elkin belt was by Lewis J. Hash,
who prepared this chapter.
The zone of sillimanite-bearing rock designated as the "Cliffside-Elkin belt" occurs in the Upper Pied-
mont of North Carolina, and extends in a northeasterly direction from the South Carolina line to near Elkin,
North Carolina. The belt is very irregular, being dis-continuous over a width of approximately 12 miles
near the South Carolina line, varying from three to five miles wide in Caldwell County, and breaking up into
long narrow bands in Alexander County. The zone traverses parts of Rutherford, Cleveland, Burke, Cald-
well, Alexander and Wilkes counties and is found in small areas of Lincoln, Catawba and Iredell counties.
This region has a rather dense population with the major portion of the population engaged in textile
manufacture and agriculture. Larger towns located within or adjacent to the belt include Shelby, Forest
City, Morganton, Valdese, North Wilkesboro, Granite Falls, and Taylorsville.
Transportation facilities in this region are adequate. Numerous state and federal highways traverse
the belt in an east- west direction and several highways and improved secondary roads parallel and cross the
4 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
belt in a north-south direction, this providing adequate roads for truck transportation. The belt is also
traversed in its southern portion by the Clinchfield and Ohio Railroad and the Southern Railroad, the middle
portion by the Southern Railroad, and the northern portion by a branch of the Southern Railroad.
Most of the deposits are within one mile of adequate water for plant operations. Major streams from
south to north include the Broad, Catawba and Yadkin rivers, all of which traverse the belt in a more or
less east-west direction. Large tributary streams flow within and parallel to the belt.
Climatic conditions are favorable for "year-round" mining with minimum temperatures seldom less
than 0° F. and maximum temperatures seldom exceeding 100° F. The mean annual temperature is approx-
imately 65° F. Rainfall is distributed rather uniformly throughout the year, precipitation ranging be-
tween 50 and 60 inches per annum.
The prevailing topography in this region is typical of the Upper Piedmont, consisting of rolling hills
and valleys. Two small mountain ranges traverse the belt, thus locally breaking the areal topography. South
of Morganton the South Mountains, reaching a maximum elevation of approximately 2,900 feet, cross the
sillimanite zone in an east-west direction. Northeast of Morganton and east of Lenoir the Brushy Moun-
tains, rising to a maximum elevation of approximately 2,500 feet, occur along the northwestern portion of
the belt. In general no association is apparent between rock types and topography except on a local scale.
Many places were found where sillimanite schist and granite gneiss traversed the mountains without either
type of rock causing noticeable topographic features. However in some areas of major concentration of sil-
limanite, the sillimanite-bearing rocks have influenced local topographic expression. An outstanding exam-
ple of this condition is Cages Mountain, north of Rutherford College, where a conspicuous ridge is underlain
by sillimanite-bearing rock.
Natural exposures of sillimanite rock in the area are rare and are confined primarily to steep bluffs
and banks of streams. Where excavations have been made for highway construction, road bank exposures
of weathered material can be found. The scarcity of outcrops is an obstacle to detailed mapping; however,
a considerable quantity of float material ranging in size from small nodules up to fragments one foot across,
usually occurs in areas of sillimanite-bearing rocks. At a number of deposits the only method of sampling
was to obtain float material. Prospecting will be required in most deposits in order to study the structure,
size, composition, etc.
The better sillimanite deposits occur predominantly in low, long, relatively narrow ridges having round-
ed tops and gentle slopes. This type of ridge forms although the sillimanite is resistant to mechanical and
chemical weathering. However, due to its fractured character steep ridges with bold outcrops do not form.
An exception is the Smith-Cliff deposit where a stream cut away the toe of the hill, leaving prominent ex-
The sillimanite-bearing rocks in Rutherford and Cleveland counties are confined to a zone about 12
miles wide. Northeastward the zone is narrower, being approximately 5 miles wide in the vicinity of South
Mountains in Burke County. It has relatively regular boundaries in this area and the individual bands of
sillimanite-containing rock are wider and more closely spaced. From the Catawba River northeastward
to Dudley Shoals, the zone is composed almost completely of sillimanite schist with only occasional small
lenses of granite or granite gneiss.
From Dudley Shoals northeastward the zone widens for a few miles and then breaks up into various
bands which decrease in width rather uniformly northeastward into Wilkes and Iredell counties. North-
east of the point where the zone branches into individual bands, the schist is generally uniform in trend but
contains a relatively low percentage of sillimanite. In this area it is possible to trace the individual schist
bands with little difficulty.
The original reconnaissance survey 3 on sillimanite in this zone indicated that the sillimanite-bearing
rock occurred in a shear zone. Field evidence collected during the current survey substantiates this obser-
vation. Individual shear zones vary up to several hundred feet in thickness and are readily evident by
their fragmentary appearance. Evidence of shearing is present in many orders of magnitude, varying from
minute intragranular shearing, seen only with the microscope, to large persistent fault planes which may
SlLLIMANITE DEPOSITS OP NORTH CAROLINA 5
be traced for considerable distances along the outcrop. The commonest evidence of shearing is found in
small fragments which have been dragged into somewhat sigmoidal shapes. The entire zone of sillimanite-
bearing rock is highly fractured and contains extensively folded schists. Small thrust faults and closely
spaced joints are numerous. In addition, local normal faults with small displacement occur.
In general, the individual bands of sheared sillimanite-bearing rock cut diagonally across the major
belt, which trends from N 40° to N 60° E. These bands strike from N 20° to N 40° and dip about 20° SE,
however, locally the dip steepens and may be as great as 50° SE. It is thought that local variations in trend
are due to post shear folding which has warped the individual schist bands. Local northwest strikes were
observed, especially in the Hollis-Casar area and northeast of Dudley Shoals.
The sillimanite schist is so thoroughly fractured that it breaks into elliptical shaped fragments or nod-
ules ("buttons") which range up to 3 inches in length. These small nodules cover the surface at many places
and are helpful in identifying and outlining the sillimanite-bearing zone, especially in areas where out-
crops are rare.
A number of different rock types occur throughout the Cliffside-Elkin belt but, in general, the prin-
cipal rock types are granite, granite gneiss, and quartz-biotite schist.
The rock in which sillimanite occurs is predominantly a quartz-biotite-sericite-sillimanite schist. Ac-
cessory minerals include garnet, graphite, chlorite, alum, pyrite, zircon, ilmenite, barite, and rutile. Near
the surface biotite and pyrite have been oxidized and have stained the sillimanite rock. Sericite occurs in
considerable quantity, locally composing approximately 50 percent of the schist. Pyrite, occurring in small
amounts throughout the zone, was the only sulphide observed. It appears to be one of the late minerals.
The other accessory minerals occur only in trace amounts, some of which apparently were original constitu-
ents. Some of these minerals were formed by metamorphism, and others by igneous solutions.
Small pegmatites and narrow quartz veins occur throughout the zone. The pegmatites, in general, aver-
age less than a foot in thickness and are rarely more than 4 feet long. The principal minerals in the pegma-
tites are quartz, feldspar, mica, and small amounts of tourmaline. Locally pegmatites compose as much as
one-half of the sillimanite-bearing rock.
Hunter and White 3 who examined thin sections of the sillimanite-bearing rock described it as follows :
"Under the microscope the sillimanite-bearing rock is seen to be a quartz-biotite schist with very minor
amounts of a plagioclase which appears to be albite. The bulk of the rock is quartz, which like the feldspar,
seems to have been a component of the original. Both these minerals are allotriomorphic. Small euhedra
of biotite are disseminated quite generally throughout the rock and locally garnets appear. Both these min-
erals probably represent original components of the rock.
"The sillimanite appears in needlelike crystals which replace the quartz, biotite, muscovite, and garnet.
The crystals occur singly or in groups which take on several distinct patterns of aggregation. Sometimes
they are in subparallel growth but following a sinuous line apparently determined by the previous position
of the linear aggregates of muscovite crystals which they tend to replace, largely with the longest direction
of the sillimanite parallel to the cleavage of the muscovite. Again, the bundles of sillimanite needles may
be heterogeneously oriented in felted aggregates, or thinly disseminated in random orientation throughout
the quartz. Of most importance, however, is the tendency for them to form bundles in parallel or sheaf-
like growth to the virtual exclusion of all other minerals. Such bundles are usually small and sparsely dis-
tributed through the rock, but in some of the more favorable localities such as Smith Cliff on Prospect
Ridge in Burke County they are larger and form a significant fraction of the rock, offering potentialities for
Granites occur on the southeast and northwest sides of the shear zone. These rocks show evidence of
some deformation but not nearly so pronounced as the sillimanite-bearing rock. Two different granites occur
throughout most of the length of the belt, both of which may be present at any one locality. These granites
appear to be of different age and one at least is apparently post shear in age. No attempt was made to map
the extent of either the granites and only a megascopic description is available. The most abundant gran-
ite contains a very high percentage of quartz and a small amount of mica and feldspar. It is fine-grained and,
as a rule, makes poor outcrops. The other granite is coarse-grained, containing a rather high percentage
6 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
of microcline feldspar and relatively low percentage of mica and quartz. This granite occurs most fre-
quently on the northwest side of the belt.
In the Dudley Shoals section an irregularly shaped area is underlain by the microcline granite, tongues
of which protrude almost across the shear zone. At this locality the granite apparently is post shear in
age and may be responsible for the higher grade sillimanite deposits which occur in the Dudley Shoals area.
The character of the parent rock of the sillimanite schist is not known. Field evidence, however, indi-
cates that the original material from which the schist was formed was of sedimentary origin, that it con-
sisted of quartzite interbedded with shale, and that the schistose layers now represent the metamorphosed
equivalent of the original shaly facies.
DESCRIPTIONS OF DEPOSITS
The following descriptions of individual deposits in the Cliffside-Elkin belt are in order from south to
north. Considerable variance of sillimanite content and local conditions of geology are evident throughout
the belt, and in the descriptions of individual occurrences these local variations are described.
The Polkville deposit is located in Cleveland County, one mile N 30° E of Polkville, in a ridge about .2
of a mile from a secondary road (Map location 6, Plate 1). It can be reached by following the secondary
road one mile northeast of Polkville. The deposit is on the north side of the road.
Considerable quartz-sillimanite schist and float pegmatite occur along the crest of the ridge, but no out-
crops were found in which to study their relationship. Float schist sampled from the ridge contained 24.8
percent sinks* (Laboratory sample 713-8). The sample was representative of the schist but did not include
any pegmatite material, which, judging from the amount of float, probably makes up at least V3 of the rock.
The richer schist may occur in narrow bands. Other samples of sillimanite-bearing rock collected from this
locality and analyzed for sillimanite content included samples 713-12 to 14.
Wards Creek Deposit
This deposit, located in Cleveland County 5.2 airline miles N 55° E of Hollis and 3.9 airline miles N 8°
W of Polkville, can best be reached by following highway No. 10 two miles north of Polkville, turning left
and following secondary road .5 of a mile, turning right and following secondary road 2.2 miles (Map loca-
tion 9, Plate 1). The deposit is in a ridge on the northwest side of the road.
Considerable float, both small nodules and schist, is found along the ridge. A few large nodules, similar
to the Dudley Shoals type, were found along the ridge. Representative samples of float material taken
along the ridge contained 33.2 percent sinks in heavy liquid concentration (Laboratory sample 713-2).
Microscopic examination revealed the presence of sillimanite, barite, rutile, ilmenite, hematite, and zircon,
with sillimanite constituting about 85 to 90 percent of the sinks. About 6 percent of the sinks was barite.
Very little sericite alteration was observed.
The deposit is believed to offer the best possibilities for development of any deposit found in this area;
however, due to the absence of outcrops only little detailed information could be obtained. The float may
be representative of the underlying schist or may have been formed from rich pockets or small zones that
was more resistant to weathering than the other material. Other samples studied from this locality include
713-1, 3 to 7.
Two sillimanite occurrences have been noted in the vicinity of Casar. One is located 3.5 airline miles
N 78° E of Casar and 3.7 airline miles N 39° W of Belwood. It can best be reached by following a paved
highway 3.7 miles northwest from Belwood, turning left and following a secondary road 3.4 of a mile (Map
location 2, Plate 1). The deposit is located y% of a mile northwest of the road near several old mica mines.
The other deposit is located one mile northwest of this locality in a large ridge adjacent to the road.
*See Laboratory results.
^ » -S
'x ^ 5
o o ^)
li. a. <r>
J c, Ji
8 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
The first deposit occurs adjacent to heavily iron stained kyanite schist and near several sheet mica peg-
matites. Because of poor exposures and insufficient time, the geological relationship of the several rock
types was not studied in detail. However the occurrence is an unusual one since no appreciable amount of
kyanite was found at any other locality in the Cliffside-Elkin sillimanite zone.
The sillimanite schist from both deposits is similar, being composed of very fine needles of sillimanite
(fibrolite). The other minerals are predominantly quartz and sericite. The amount of iron stain appeared
to be relatively low. Microscopic examination (Laboratory sample 713-10) of sinks in heavy liquid revealed
the presence of sillimanite, ilmenite, barite, staurolite, rutile, zircon, and a few zircon inclusions. About 95
percent of the sinks was sillimanite with practically no alteration to sericite. Most of the remaining heavy
mineral content was ilmenite.
Samples taken from outcrops and float schist from the two deposits contained 16.0 (Laboratory sample
713-10) and 16.4 percent (Laboratory sample 713-11) sinks, the former being from the deposit near the mica
mines. The sample is believed to be representative of the schist, but does not include any material from the
numerous small pegmatites that are found throughout the schist. It is estimated that the pegmatites make
up from Vi to % of the material in these deposits (See also laboratory sample 648-C).
Smith Cliff Deposit
The Smith Cliff deposit is located in Burke County, seven miles (airline) S 42° E of Morganton and can
be reached by following North Carolina Highway No. 18 for 8.5 miles southeast of Morganton, turning left
and following a country road one mile to the east (Map location 8, Plate 1). The cliff is about 14 °f a m ^ e
to the north, adjacent to and on the north side of Henry Fork Creek.
The sillimanite at Smith Cliff occurs in a quartz-biotite-muscovite-sericite-sillimanite schist, quartz and
biotite being the most prominent minerals in the unweathered schist. Tourmaline, graphite, garnet, feld-
spar, and pyrite occur in small quantities.
The schist forms a nearly continuous outcrop for 600 to 700 feet along the face of a very steep cliff (Plate
2). The cliff, which varies from 300 to 330 feet high, is very conspicuous and differs from the long narrow
type ridge which generally underlies sillimanite deposits in the belt.
There is a relatively large fold (mostly obscured by overburden) in the eastern part of the deposit, the
axis of which extends up a large draw (Plate 2). A basic dike occurs in the draw near the fold axis. The
schist west of the fold axis has a general strike of from N 30° E to N 60° E and east of the axis strikes from
N 30° W to N 60° W. Small gentle folds are found throughout the cliff, but no area was observed in which
intensive crinkling occurred. West of the fold axis and including the major portion of the deposit, the rocks
dip from 15° to 25° NW into the ridge, while the rocks in the eastern part dip northeastwardly into the
ridge at about the same angle. Much of the schist had two planes of schistosity, one being roughly parallel
to the dip of the beds, the other varying at angles up to 45°.
Joint surfaces are very prominent throughout the cliff, some being from 30 to 35 feet across. There are
two major joint systems — one striking from N 15° to 20° W and the other striking from N 40° to 45° E;
other minor joint systems occur. All joints have near vertical dips.
A characteristic of the deposit is the great amount of igneous activity. This activity is represented by
granite, small pegmatites and granitized schist; gradation from schist to granite is common. Small lenses
of granite seldom exceeding 25 feet in width and 100 feet in length are found at numerous places through-
out the deposit.
Small pegmatites, from 3 inches to over 3 feet wide and several feet long, occur as lit par lit injections.
These pegmatites, in places, compose from 1/5 to 1/3 of the rock. The principal minerals in the pegmaties
are quartz, feldspar, mica, and small quantities of tourmaline. The mica books range up to 3 inches in
The granitized schist, which approaches a granite gneiss in both mineral composition and structure, is
found in many places, often forming near-vertical ledges that extend for several hundred feet along the
face of the cliff. This rock appears to have been a schist that was saturated and altered by invading igne-
10 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Lenses of quartzite seldom over three feet wide and four or five feet long occur in many places over
the face of the cliff. Some of the quartzite appears to have been completely replaced by granite. Other
quartzite shows evidence of granite having partially invaded and replaced it. No sillimanite was observed
in the quartzite. The rock along the top of the ridge is highly weathered and iron stained, but in places,
100 feet or more down the cliff, relatively unweathered material occurs.
The sillimanite in this deposit occurs in two forms — as nodules or bundles up to two inches in diameter,
which are composed of many small crystals, and as individual sillimanite crystals disseminated through the
schist. The nodules are conspicuous in and near the prospected area (Plate 2) and as float along the top
of the ridge. These nodules are very resistant to weathering, consequently standing out and giving the
rock a "knotty" appearance. Some are practically pure sillimanite, while others contain sericite interlocked
with the sillimanite grains and are badly iron stained.
The largest quantity of sillimanite occurs as small individual sillimanite crystals, disseminated through-
out the schist. The sillimanite content of the schist ranges up to 11 percent (Laboratory samples 660-1 to
8). Locally, areas are found in which the sillimanite content is above 11 percent, but none contained a min-
able quantity of ore. The rich areas occur along fracture zones. Most of the schist sampled contained from
five to seven percent sillimanite; however, two samples from different localities on the cliff, representing a
substantial quantity of ore, contained 9.2 and 11 percent. Considerable sericite occurs with the sillimanite
in this deposit.
Flotation work at the North Carolina State College Minerals Research Laboratory in Asheville showed
that a good sillimanite product can be prepared from this ore, but recoveries were low. It may be possible
to produce some coarse sillimanite ( + V4 inch material) from the nodules, but it is believed this will be very
difficult, if not impossible, when hard rock is encountered. Any coarse material produced from the weath-
ered schist probably will be iron-stained and will contain some sericite interlocked with the sillimanite
Saw Mills Deposit
This deposit is located in Caldwell County 3.3 airline miles N 30° E of Rutherford College and 2.7 air-
line miles S 35° W of Saw Mills. It can be reached by following the highway from Rutherford College one
mile north of the Catawba River, turning right and following the secondary road IV2 miles (Map location
7, Plates 1 and 3). The deposit is in the ridge east of a creek.
Large nodules similar to the Dudley Shoals type can be found over a considerable area, but they do not
exist in as large a quantity as those at Dudley Shoals. It is thought that the nodules occur more or less at
random throughout the schist. A sample of schist float taken at numerous places along the ridge contained
9.4 percent sillimanite (Laboratory sample 686).
Cages Mountain Deposit
The Cages Mountain deposit is located in Caldwell County, 6 miles (airline) S 25° W of Hudson. It is
well exposed in a deep road cut on the highway from Whitnel to Rutherford College (Map location 1, Plates
1 and 3). The property, owned by Silvio Martinat of Lenoir, occurs on a typical long, relatively narrow,
"sillimanite ridge" having a rounded top and gentle slopes. This ridge is the most outstanding topographic
feature in the area and can be seen above the other ridges in the area from a distance of 15 miles.
The rock in which sillimanite occurs is predominantly a quartz-biotite-sericite-sillimanite schist. Other
minerals which occur in small amounts include ilmenite, pyrite, limonite, rutile, zircon, barite, and alum.
Quartz, biotite, and sericite are the most abundant minerals in the unweathered schist. Near the surface
biotite is thoroughly weathered; weathering having produced considerable amounts of iron oxide to depths
of from 20 to 30 feet. This iron oxide has stained the sillimanite appreciably. Sericite occurs in both weath-
ered and unweathered rock. A large percentage of the sericite has formed by the retrogressive alteration
of sillimanite. Alum occurs as surface coatings in the rock near the bottom of the road cut but is not found
in weathered material near the surface. This mineral was formed by the reaction of acid, resulting from
the oxidation of pyrite, upon aluminous materials. Ilmenite, zircon, barite and rutile are either of hydro-
thermal origin or represent detrital material in the original sediment.
— 1_' L.
12 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
The schist has a general east-west strike but local variations from N 15° W to N 55° W occur. It gen-
erally dips from 15 to 30° to the south, however, where intensively folded, dips to the north occur. The
most intensively folded area is immediately south of the quartzite lenses (Plate 4). Both north and south
of the quartzite zone, the folds become broader, with limbs dipping approximately 30°.
Two joint systems are conspicuous in the road cut; the individual joints strike N 25° W and N 45° W,
having dips of approximately 80° SW. The joints are spaced at intervals of 2 inches to 3 feet. Many joint
surfaces 4 to 5 feet across are exposed. In addition to the joint systems and folds a normal fault striking
parallel to the ridge and dipping nearly vertical is exposed in the road cut. This fault has a small amount
of gouge in the zone of movement, however, it is not thought that the displacement is large, thus it has little
significance in the geology of the deposit.
In addition to the sillimanite-bearing schist, other exposed rock types include quartzite, pseudo-diorite,
quartz veins, and pegmatites (plate 4). The quartzite is a relatively massive and conspicuous band in the
road cut. Pseudo-diorite occurs as lenses in the schist adjacent to the quartzite and probably represents the
replacement of originally limey lenses by siliceous material. Quartz veins and pegmatites are numerous,
ranging up to one foot in width and 5 to 6 feet in length. Small pegmatites, approximately one inch in
width have invaded the schist both as lit par lit and as crosscutting bands. The pegmatites and quartz veins
are more numerous in the intensively folded zone.
Except in the road cut the only exposures of bed rock are along a small valley near the east end of the
ridge. These outcrops are narrow and highly weathered, revealing little geologic information. Consider-
able sillimanite schist float occurs along the ridge, fragments varying in size from two or three inches to a
foot or more in diameter. Firm rock should be encountered throughout most of the ridge at depths from
10 to 15 feet.
The sillimanite in this deposit occurs as small nodules up to V4 mcn m length and as individual crystals
disseminated throughout the schist. The nodules are composed of many small interlocked sillimanite crys-
tals, so that only fine concentrates can be produced from this ore.
The sillimanite content of the schist varies from 5 to 14.3 percent (Plate 4) (Laboratory samples 648-D
and H, 668-B to F, 669-1 to 9) . Microscopic examination of the concentrates made from the ore reveals that
up to 60 percent of the sillimanite crystals are partially altered to sericite. The schist with the highest
sillimanite content is immediately south of the quartzite body at approximately the center of the road cut.
A zone 80 feet wide contains 10 to 15 percent sillimanite. It is noteworthy that the most intensively folded
and intruded zone contains the highest percentage of sillimanite. The sillimanite content decreases rather
uniformly on both sides of this rich zone. Samples of float sillimanite schist taken at random along the
crest of the ridge over an area 50 feet wide and 400 feet long varied in sillimanite content from 11.5 to 14
percent. Float samples taken from above the road cut on either side of the road contained 8.2 and 7 per-
cent, respectively, while samples of bed rock taken at the bottom of the road cut directly below at a depth
of approximately 35 feet contained from 10.2 to 14.5 percent sillimanite. Since sillimanite is a mineral
very resistant to chemical weathering, the reasons for this variation in sillimanite content is not clear, but
should this same proportion hold true for the crest of the ridge, the schist underlying it should contain
above 20 percent sillimanite. Some prospecting will be necessary to determine accurately the percentage,
quality, and quantity of sillimanite near the crest of the ridge.
Because of the mixed sillimanite and sericite and the difficulty of separating the two minerals, a pure
sillimanite concentrate has not been prepared from this ore. If a mixture of sillimanite and sericite can be
used by the ceramic or refractory industries, an ample tonnage of ore can be obtained from this deposit.
Dudley Shoals No. I Deposit
The Dudley Shoals No. I deposit is located in Caldwell County, four and one-half miles airline N 21° E
of Granite Falls, and one and one-half miles airline S 79° W of Dudley Shoals (Map location 3, Plates 1 and
3). It can be reached from Dudley Shoals by following the secondary road leading to Cedar Valley one
mile west, turning left and following a secondary road for one-half mile. The deposit is located south of the
road on the property of Gather Teague, Dudley Shoals.
S Si s> g &
5 "\i ^i sj o>
5»:S 5- 5- 3
14 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Little detailed structural information can be obtained from the deposit because of the lack of outcrops.
The sillimanite occurs in a quartz-sericite-sillinianite schist, which contains some biotite in the unweathered
zone. Microscopic examination of sinks in heavy liquid concentration tests of samples from this deposit
revealed the presence of sillimanite, specular hematite and two to three percent staurolite. Sillimanite altera-
tion to sericite apparently is less than at any other deposit examined, but a considerable quantity of sericite
is found locally in the deposit.
The outcrops examined are in a road cut, about Vs °f a m il e eas t of the center of the deposit, and a
small outcrop northeast of the crest of the ridge. Rock exposures in the road bank have a strike of from N
5° E to N 20° E and dips from 35° to 40° SE. The schist on the ridge strikes N 20° E and dips 40° SE. The
rock appears to be folded near the crest of the ridge, the western part of the deposit striking roughly N 70°
W and dipping to the southwest (Plate 5) . The trend of the ridge changes from southwest to northwest and
considerable float, containing a relatively high percentage of sillimanite, is found along and on the north
side of the ridge.
Considerable sillimanite schist float is found along the crest of the ridge, the largest concentration be-
ing near its highest point. The float ranges from pieces one inch to three feet in diameter. Firm rock should
be encountered at a depth of approximately 20 feet.
Sillimanite-bearing schist in the Dudley Shoals area is bounded on the west, north and east by granite.
Evidence which was obtained from the deposit suggests that the sillimanite occurs in an anticline fold which
plunges to the north under the granite. The granite apparently invaded the schist parallel to the schistos-
ity, completely assimilating the rock above and on both flanks of the folds. Subsequently, erosion has
removed part of the granite, thus exposing the schist.
This deposit is unique in that the sillimanite occurs in so many different forms, varying from small
needle-like crystals disseminated through the schist to large nodules, up to two feet in diameter that are prac-
tically pure sillimanite.
Most of the sillimanite occurs as small crystals disseminated through the schist, but a substantial quan-
tity of coarser material is also present. The schist contains nodules up to two inches in length and one inch
in width which are composed of many small interlocked sillimanite crystals. Locally, these nodules compose
from 40 to 50 percent of the schist. No evidence could be found to indicate the frequency of occurrence of the
large nodules (up to one foot across). However, from the number occurring as float, especially in the soil
along the northeast slope of the ridge, it is believed that they compose a substantial part of the ore, probably
occurring close together in fracture zones. If these nodules occur in sufficient quantities they can be re-
covered as coarse sillimanite ; otherwise they will increase the grade of ore. Information about their occur-
rence can be determined only after prospecting ; however, considering both the small and large nodules, this
deposit is believed to offer better possibilities for recovering coarse material than any other deposit exam-
ined in the Cliffside-Elkin belt.
The schist float occurring along the top of the ridge contains from 16.5 to 31.8 percent sillimanite (Plate
5). These samples were taken close together at the highest point on the ridge. There are about equal
amounts of both types of material in the immediate area. Samples taken northeast of the ridge crest con-
tained 18.3 and 20.8 percent sillimanite (Laboratory sample 707-A and B).
In addition to the schist, there is an area north of the ridge where sillimanite has been concentrated in
the soil. The sillimanite in the soil ranges from fragments of small crystals ( — 325 mesh) to nodules two
feet or more across. Surface samples to a depth of six inches had a sillimanite content up to 18.0 percent
(Plate 5), about half being minus 16 mesh material. Other samples taken at depths from six inches to five
feet contained up to 6.1 percent sillimanite (Samples 730-732-E, and 758).
Dudley Shoals II
Sillimanite schist occurs in the ridge .5 of a mile northeast of the deposit described in detail (Dudley
Shoals I, Map location 3, Plates 1 and 3). Large and small nodules, made up of many individual sillimanite
crystals, and minus 16 mesh sillimanite crystals occur in the soil. The schist does not crop out on the ridge
and only a small amount of schist float occurs, so that additional exploration work will be necessary to accu-
rately appraise the deposit.
C f= Si
- E o (/•
— — Z
st Lean in S
st Rich in Sil
■£ Z £ *
u u fc C
16 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Large nodules, up to three feet across and weighing two or three hundred pounds, are present. Num-
erous rockpiles found along the edges of fields contain 50 percent or more large sillimanite nodules. The
nodules are found throughout an area extending for approximately one-half mile southeastward from the
road. Several tons of nodules could be collected from this area.
In addition to the large nodules, small nodules and fine sillimanite have been concentrated in the soil
by weathering and erosion. A surface sample, taken over an area of approximately 50 feet by 200 feet,
contained 10.7 percent sinks. This deposit has not baen sampled in detail, but it appears to contain a con-
centration of sillimanite in the soil all along the ridge which is approximately V4 of a mile long and several
hundred feet wide (Samples 708-A and 732-F).
Dudley Shoals No. Ill
Another promising deposit in the Dudley Shoals area is located approximately 1/3 of a mile southwest
of the Dudley Shoals No. II deposit (Map location 3, Plates 1 and 3). The deposit occurs on the south side
of a small stream in a relatively steep-sided ridge.
There is considerable sillimanite schist float along the ridge, which is covered by timber, but no large
nodules occur. Samples of schist float taken at random along the ridge contained 20.8 percent sinks. (Lab-
oratory Sample 708-B). Microscopic examination of sinks in heavy liquid concentration revealed the pres-
ence of sillimanite, hematite, staurolite, small amount of garnet, and zircon. Sillimanite, about 5 to 7 percent
being altered to sericite, made up about 85 percent of the sinks, and was mostly needle-like. Nine to ten
percent of the sinks was hematite and 2 to 3 percent was staurolite (See also samples 707-C and D).
Because of the higher sericite alteration and absence of coarse material this deposit does not appear to
offer as great commercial possibilities as do the other two deposits in this area; however, it is noteworthy
because of its high sillimanite content.
Ellendale School Area
The Ellendale School area is located in Alexander County, six miles west of Taylorsville, adjacent to
highway 90 (Map location 4, Plate 1).
Several bands of schist occur in this area, notably 1/10 of a mile east of Ellendale School and extending
for three or four miles to the north. These bands, in general, are from 50 to 100 feet wide, strike approxi-
mately N 30° E and dip from 20 to 40 degrees to the southeast.
Both fine needles and nodules of sillimanite occur disseminated throughout the schist. Microscopic ex-
amination of concentrates from these deposits show that sericitization has been extensive. In places it ap-
pears that sillimanite once composed as much as 50 percent of the rock, but that later solutions altered a large
part of it to sericite.
Fox's Orchard Deposit
The Fox's Orchard deposit is located in Iredell County, 8 airline miles N 37° E of Hiddenite and 4 air-
line miles N 32° E of Smith's Store. It can be reached by following highway 115 four miles south from the
Wilkes County line, and following the secondary road west to the Alexander County line (Map location 5,
Plate 1). This deposit is located in the ridge south of the road.
Two types of sillimanite occur in this area. Nodules, composed of fine-grained sillimanite and sericite up
to two inches in diameter, occur along the northeastward trending ridge. These nodules occur as zones in a
quartz gneiss. The nodule-bearing zones, which average perhaps fifteen feet in width, strike to the northwest.
Zones are thickly spaced on the southwest end of the ridge, being approximately 100 feet apart in places,
with the space widening progressively to the northeast. Locally, the nodules compose as much as 50 percent
of the rock, standing out in relief, thus imparting a knotty appearance to the rock. Microscopic examina-
tion and laboratory concentration tests have shown that this material is badly iron-stained, and that sericite
alteration has developed to such an extent that a commercial product can not be made from the ore. (See
Laboratory samples 648-E, 662-1 and 2).
The other type of sillimanite found in this area consists of interlocking needles associated with large gar-
nets. The principal zone of this material is about 50 feet wide, approximately 300 feet long, and is located
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 17
on the north side of Rocky Creek. The sillimanite content varies considerably. It is very fine-grained and
appears to have no commercial value.
The sillimanite schist in this area, the northeastern part of the belt, consists of long, narrow bands
striking from N 30° E to N 80° E. Although locally the' strike of the schistosity varies considerably, strik-
ing northwest in many places, the individual bands are generally uniform in both width and trend. The
bands vary from a few feet to several hundred feet in width, and are usually from one to three miles apart ;
the most abundant country rocks are granite and granite gneiss.
Most of the samples from this area were collected from road cuts. The sillimanite content of individual
bands appeared relatively uniform over considerable areas. The locations and sillimanite contents of some
of the more promising deposits are as follows :
A deposit, located in Wilkes County and outcropping in the road bank on Highway 16 on northern slope
of the Brushy Mountains, is about 50 feet wide and can be traced easily through the mountain by the float
schist. A sample taken from the road cut contained 12.7 percent sillimanite and a sample of float schist
from the ridge northeast of the road contained 12.2 percent sillimanite (Laboratory samples 696-A and B).
A deposit, located in Alexander County and outcropping in the road bank on highway 16, four miles
south of Taylors ville, contained 11.0 percent sillimanite. A sample of high grade float from this deposit
contained 17.7 percent sillimanite (Laboratory samples 694 and 695) .
Numerous other samples taken from the Wilkesboro-Taylorsville area contained under 10 percent silli-
manite. It is probable that better deposits in this area will be found; however, geological conditions indi-
cate that the chances of finding commercial deposits are better in other sections of the zone. Sericitization
has been very extensive in the Wilkesboro-Taylorsville area.
SILLIMANITE IN THE WARNE-SYLVA BELT
The following descriptions of the sillimanite occurrences in the zone between Warne and Sylva, North
Carolina, were prepared by Earl C. Van Horn.
The sillimanite-bearing rock designated as the "Warne-Sylva belt" occurs in a more or less continuous
zone trending northeastwardly from the Georgia-North Carolina line, across Clay and Macon counties to
the vicinity of Sylva in Jackson County. The major population in this region is concentrated in the more
fertile valleys. Towns adjacent to the belt include Hayesville, Franklin, Dillsboro, and Sylva.
U. S. Highway 64 traverses the Warne-Sylva belt in the vicinity of Hayesville in Clay County. State
highway No. 28 crosses the belt northwest of Frankl'n, and U. S. Highway 23 crosses the belt east of Dills-
boro. Secondary roads, adequate for truck traffic, provide access to most of the area. A branch line of
the Southern Railroad extends to Hayesville from Andrews, North Carolina. Franklin is provided rail
transportation by the Tallulah Falls Railroad which joins the main line of the Southern at Cornelia, Georgia.
The northern end of the belt, in the vicinity of Dillsboro and Sylva, is served by the Asheville-Andrews-
Murphy Branch of the Southern Railroad.
Major streams which occur in the area and cross the sillimanite-bearing zone include, from south to
north, the Hiwassee, Nantahala, Little Tennessee, and Tuckasegee Rivers. Tributaries of these streams pro-
vide adequate water to most of the deposits.
Climatic conditions of the Warne-Sylva area are favorable for mining during most of the year, how-
ever, during the winter light snows and extended periods of rainy weather would interfere with open cut
operations. Temperatures range between minus 5° F. to approximately 95° F. with an average tempera-
ture of approximately 60° F. Rainfall is heavier than in the Piedmont area, averaging approximately 60
inches per year.
North Carolina State Library
jlaleigh, N, Q,
18 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Major mountain ranges which traverse or parallel this sillimanite belt include Davy and Cherry Moun-
tains, which reach elevations of approximately 2,900 feet; Nantahala Mountains which have elevations of
approximately 5,000 feet ; and Cowee Mountains which reach elevations of approximately 5,000 feet. Asso-
ciation between topography and the sillimanite-bearing rock is not evident in the field.
Sillimanite is known to occur in two slightly different settings in western North Carolina. The most
important sillimanite occurrences are in a well defined belt, up to 4,000 feet wide, extending from near Warne,
in Clay County, northeastwardly through Clay and Macon Counties to Sylva in Jackson County. The
occurrences between Warne and Sylva are referred to in this report as the "Warne-Sylva sillimanite belt."
Of apparently lesser importance are erratic occurrences of sillimanite near Grassy Ridge and Big Ridge
mica mines south and southwest of Waynesville in Jackson and Haywood Counties, on the north tip of Rip-
shin Mountain near the headwaters of South Hominy Creek in Buncombe County, and in the Mars Hill area
of Madison County. Sillimanite of rather low concentration occurs in these isolated localities in a garnet-
biotite gneiss. The host rock is moderately massive but a pronounced schistosity usually is present.
The belted alignment of the Warne-Sylva sillimanite belt is at only slight variance with the regional
geologic mapping of Keith 5 (1907), since it parallels roughly the formational trends which Keith showed.
In certain localities, particularly in the western part of Macon County, the sillimanite belt does cross several
of Keith's boundaries. The geology of the region, however, is characterized by transition and lithologic sim-
ilarity, so that slight difference of interpretation and classification are of little consequence. The nearly
straight trend of the sillimanite belt, even when confined to a single formation, is somewhat unusual for the
region, and slip cleavage control in a fault zone is not improbable.
The Warne-Sylva sillimanite belt is mostly a part of the series which Arthur Keith mapped as Carolina
gneiss 4 (1901). Locally this rock includes quartz-mica schist and gneiss, garnet-mica schist, gametic and
hornblendic quartz gneisses, "graywacke," and banded granite gneisses. Any of these rock types may be
graphitic in places. Pegmatite material may be present as migmatite and as dikes, veins, and pods. The
mineral content of the sillimanite-bearing rock includes muscovite, hydromica, biotite, chlorite, hornblende,
quartz, feldspar, kyanite, sillimanite, staurolite, garnet, magnetite, ilmenite, rutile, titanite, brown tourma-
line, clinozoisite, epidote, zircon, and barite. The rock has been metamorphosed extensively, and is charac-
terized by well developed schistosity and crenulations, and major and minor intra-formational faults.
In the Hayesville area, immediately northwest of the sillimanite belt, is a very old granite complex which
Keith 5 mapped as Archean granite. This complex possibly was a biotite granite which has been metamor-
phosed to granite gneiss and feldspathic schist and in many places is now indistinguishable from the rock
mapped by Keith as Carolina gneiss. At some distance to the southeast of the sillimanite belt, zones of
dark hornblende gneiss occur. Locally, masses of pyroxenite and peridotite rock occur in association with
the hornblende gneisses. Sillimanite apparently is not related in any way to either the hornblende gneiss
or to the ultra-basics.
The granite complex which occurs northwest of the sillimanite belt in the Hayesville area is not present
in the Tusquitee-Burningtown area but quartzose rock of the Great Smoky formation 5 occurs along the
northwest side. Hornblendic rocks are near the southeast side of the belt. In other respects the geology
is much the same as that described in the Hayesville area.
Published data on the general geology of the Etna-Cowee area are not available. All of the rock types
seem to intergrade, thus geologic mapping on a reconnaissance scale is of little specific value to the present
studies. Locally the rocks of the area include quartz-mica gneiss and schist, granite gneiss, garnet and kya-
nite gneisses and schists, "graywacke," hornblende gneiss, pegmatite, and, in one locality, meta-amphi-
bolite. Granites occur some distance away, both to the north and south. The sillimanite belt is associated
with all but the granite, hornblende gneiss, and "graywacke." Metamorphic characteristics are little dif-
ferent from similar rocks to the southwest, there being crenulations, several ages of cleavages, and a well
developed regional schistosity. Strikes and dips are irregular locally. Average strikes are N 50° E in the
Etna vicinity and N 60° E near the Cowee Mountains.
20 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
The general geology of the Greens Creek area is a continuation of that in the Etna-Cowee area. Rocks
within the limits of the sillimanite belt are quartz-mica schist and gneiss, in places gametic, but with a larger
proportion of schistose phases than in the Etna-Cowee area. Pegmatite material is more abundant in the
schist and except for local variations biotite is less conspicuous than to the southwest. Associated minerals
in the sillimanite belt are no different than in areas to the southwest. Of rocks adjacent to the sillimanite
belt, hornblende gneiss occurs very close to the southeast side of the belt and rocks to the northwest are
more quartzitic than elsewhere.
The southwest portion of the Sylva area involves the same geologic setting as has been described pre-
viously, a single exception being the preponderance of very hard, coarse-grained granite gneiss just south-
east of the sillimanite belt. The geology of the northeastern part of the area is complicated by the presence
of the Webster ring dike, an intrusion or series of intrusions of peridotite-pyroxenite rocks, principally of
dunite. There seems to be a tendency for the sillimanite belt to veer northward tangential to the ring dike.
This tendency is reflected in the gradual change in strike from Greens Creek to Sylva. Field studies have
not revealed specific mineralogical effects of the ring dike on sillimanite other than a gradual decrease and
final disappearance of sillimanite occurrences.
The last observed sillimanite occurrence, 400 feet northeast of the northeast corner of the Sylva city
limits, is 1300 feet from the ring dike. Sillimanite-free feldspathic quartz-muscovite schist has been ob-
served beyond the sillimanite, in the vicinity of Liberty School. Considerable reconnaissance study has not
revealed additional sillimanite-bearing schist either in an extension to the northeast, within the ring dike,
or in a drag fold east of the ring dike.
It is doubtful, from present indications, that the quality of sillimanite in weathered rock would differ
greatly from that in fresher rock. It is known that the percentage of sillimanite varies along the strike
and similar changes down the dip are to be expected. Discounting mechanical concentration of sillimanite
in overburden material, any difference in the quantity of sillimanite in weathered and unweathered rock
would be attributed to normal changes down dip. Previous experience in similar rocks nearby indicates
that chemical weathering will extend to depths of several hundred feet in many places. Depths to firm
though still weathered rock will vary principally with the elevation and horizontal distance as referred to
present drainage. Except in places where large streams have caused rapid erosion, generally firm rock is
estimated to be 30 feet to 60 feet below the surface. For short distances, near larger streams, overburden
may be 5 feet to 20 feet deep, and in high areas which are well removed from active drainage, over-
burden on the order of 100 feet would not be improbable.
DESCRIPTIONS OF DEPOSITS
The descriptions of sillimanite deposits in the Warne-Sylva belt are given in the following pages. The
deposits are described in order from south to north. These deposits represent the northeastwardly contin-
uation of the sillimanite occurrences in the Davy Mountain area of Georgia described by Furcron and
Brasstown Church-Cherry Mountain Deposits (Plate 6)
Sillimanite deposits in this area have been described previously by Furcron and Teague. 2 The silli-
manite belt begins in the vicinity of Winchester Creek, two miles southwest of Brasstown Church. South-
west of Winchester Creek only kyanite occurs in the schist zone. Between Winchester Creek and Brass-
town Creek sillimanite is present in minor quantities. Northeast from Brasstown Creek the sillimanite in-
creases in importance as a mineral constituent of the quartz-muscovite schist, being present in from three
to five fairly distinct zones, each about 200 feet thick, over a belt width of about 2,000 feet. The sillimanite
in the northwest portion of the belt is of the fibrous type, and is evenly distributed throughout the schist.
The southeastern-most zone, which includes most of the Brasstown Church grounds, is intimately associated
with pegmatites and contains additional sillimanite in the form of bladed aggregates. Teague and Furcron
report a single sample from the locality as containing 4 percent sillimanite. 2
Northeastwardly across the state line into North Carolina, exposures of sillimanite are scarce for nearly
one mile, but appear once more on the slopes and brow of Davy Mountain. Considerable float sillimanite oc-
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
— z *
5 o 2 °
< M Z UJ
O 3 I-
Z uj ID O
31- < Z
aa z uj u
5 t t
.. ^3 2 zu
ui a. ^- — — -r
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 23
curs but rock in place is not sufficient to warrant sampling. From Davy Mountain to Cherry Mountain,
three and one-half miles southwest of Hayesville, zones of quartz-muscovite schist contain minor quantities
of fibrous sillimanite, and bladed aggregates are represented by float, but the deposits are not attractive as
sources of sillimanite ore.
Hyatt Mill Creek Deposits (Plate 8)
Between Cherry Mountain and the mouth of Hyatt Mill Creek, the sillimanite belt follows the northern
portion of Hyatt Mill Creek-Coleman Creek drainage along the divide separating Hyatt Mill and Blair
Creeks. Individual sillimanite zones are less definite and the belt narrows generally to less than 1,000 feet.
Outcrops of sillimanite-bearing schist are found only at wide intervals, but float material indicates contin-
uity of the deposits. One of the better exposures is on a knoll near an abandoned house on a small tributary
of Coleman Creek, 2,400 feet N 78° W of the mouth of Mattheson Cove. A zone of about 30 feet in width
contains prismatic and fibrous sillimanite, with minor quantities of bladed aggregates, in quartz muscovite-
biotite schist inter-laminated with thin bands of pegmatite. A representative chip sample (Laboratory
No. 606) ) contained 12.6 percent sinks, approximately 50 percent of the sillimanite showing sericite in-
Sillimanite schist crops out for a width of about 100 feet in the bank of a county road 500 feet east of
Mattheson Cove, but alteration to sericite seems to have been great and only 4.3 percent sillimanite was re-
ported from a sample taken in 1942 by Messrs. Calver and McDaniel.
About 3,300 feet N 50° E from Happy Top Church, 1.7 miles south of Hayesville, blades of sillimanite
are prominent along an unimproved road and in an open pasture. Occasional narrow outcrops of quartz-
muscovite schist and pegmatite were sampled by Messrs. Calver and McDaniel, and found to contain 4.2
percent to 6.0 percent sillimanite. Similar results were obtained from material of the same type, 3,300 feet
farther northeast, near the mouth of Hyatt Mill Creek.
Downing Creek Deposits (Plate 9)
Alluvial material conceals any sillimanite deposits from the Hiwassee River at the mouth of Hyatt Mill
Creek to the vicinity of Oak Forest Church, one and one-half miles southeast of Hayesville. From Oak For-
est Church the belt continues northeastward along the south side of Downing Creek to its headwaters, a
distance of three miles. Two zones of sillimanite-bearing quartz-muscovite schist, each 100 to 300 feet wide,
are separated by 200 feet to 400 feet of quartz-mica gneiss and garnet-mica gneiss. Outcrops are present
at frequent intervals, but there are no continuous sections which show the entire width of the zones. Peg-
matite is present in pods and veins, and in disseminated form. Concentrations of coarse-bladed sillimanite
are fairly abundant in talus, enough to have been collected in the past and sold as kyanite in small lots.
Boulders of nearly pure sillimanite, weighing 300 pounds or more, have been found along the upper terrace
of Downing Creek. An outlying zone of quartz-muscovite gneiss contains small amounts of sillimanite on
the north side of Downing Creek, from near Patterson Mill Creek to Pecky Wood Branch. These occur-
rences are poorly defined and of low sillimanite content (up to 3 percent).
The cemetery and a road cut at Oak Forest Church show sillimanite contents in a quartz-muscovite
schist. Prismatic and fibrous sillimanite appear to be present in quantity, but channel samples revealed
less than 6 percent sillimanite (Laboratory sample 673). A channel sample collected by Messrs. Calver and
McDaniel analyzed 9.9 percent sillimanite. Pod-shaped boulders of sillimanite, as much as 12 inches across,
are present along and across a hill between Oak Forest Church and John Reece Branch. Both bladed and
prismatic sillimanite occurs along a road leading from Oak Forest Church to John Reece Branch. Quality of
the material is similar to that nearer the church.
Sillimanite float is noticeable across-country, as well as along a secondary road from John Reece Branch
to the head of Downing Creek. Prismatic sillimanite, distributed throughout the quartz-muscovite schist,
outcrops along branches and side roads which cross the belt. In the pasture of A. D. Sellers, 2,400 feet N 75°
E of Downing Creek Church, two channel samples across 20 feet of section contained 8.4 percent sinks
(Laboratory sample 675). The rock is a gametic quartz-muscovite schist having thin pegmatite layers.
A sample 1,400 feet east of the mouth of Bob Prater Branch contained 6.2 percent sinks (Laboratory sample
604). Other samples from this area studied include 605 and 607.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 25
In spite of the seeming abundance of coarse bladed sillimanite float in the Downing Creek vicinity, it
is thought unlikely that coarse sillimanite can be found in sufficient quantity to be of important commercial
value. An exception would be for the owners to collect sillimanite nodules and boulders in small lots for
sale to a broker. Apparently the coarse sillimanite has been supplied by many small, scattered pegmatite
pods and stringers and concentrated slightly in surficial deposits.
Sap Sucker-Cold Branch Deposits
Northeast from the head of Downing Creek, the sillimanite belt crosses a small divide into Tusquitee
Creek drainage. Lean sillimanite-bearing schist occurs in a road cut at the first crossing of Sap Sucker
Branch. A better sillimanite exposure is present in a road cut 600 feet southeast of the mouth of Pecky
Wood Branch, where about forty feet of schist contains visible sillimanite. The rock is slightly graphitic
and contains thin quartzitic beds and small pegmatite stringers. A channel sample (Laboratory sample
674) of the section contained 7.6 percent sinks.
A 20-foot zone of pegmatized schist, containing disseminated fibrous sillimanite, crops out 3,000 feet
north of the mouth of Pecky Wood Branch. This occurrence could not be traced along the strike. One
thousand feet south of this location, Messrs. Calver and McDaniel sampled a 40-feet outcrop of mica schist
which was estimated to contain 2.5 percent sillimanite.
Northeast of Sunday Branch, and 6,000 feet above its mouth, four narrow zones of sillimanite-bearing
schist occur over a cross-strike distance of 1,000 feet. Kyanite predominates over sillimanite in most of
the exposures, but Messrs. Calver and McDaniel sampled one 20-foot gametic zone which contained 7 per-
Minor amounts of sillimanite occur with kyanite in graphitic and gametic schists on Cold Branch, but
these deposits have no commercial value. From Cold Branch to the head of Tusquitee Creek, only a few
exposures of the schist are present and these are devoid of visible sillimanite.
Tusquitee Gtap-Jarrett Creek Deposits
Beginning at the Tusquitee Gap road, one-half mile north of the mouth of Bluff Branch, a lean silli-
manite zone has been traced into Nantahala River drainage, crossing the divide about 200 yards east of
Tusquitee Gap. The schist contains much sericite and small amounts of kyanite and graphite. From a
point about one-half mile north of the divide, outcrops of the schist were not found, but float sillimanite
occurs on either side of a projection of the belt. Sillimanite appears in four thin zones on the Aqu one-Rain-
bow Springs Road northeast of the Horseshoe of Nantahala River. At a sharp bend in the road, on a south-
west extension of the Jarrett Knob ridge, a 30-foot zone of intensely pegmatized mica schist contains about
5 percent fibrous sillimanite and minor amounts of the coarse-bladed sillimanite. Three additional zones
of fine-grained muscovite schist occur for a distance of about one and one-quarter miles to the northeast, but
they are too thin and too lean to be of value.
On the Rainbow Springs road, 2,000 feet west from Jarrett Creek, a zone of muscovite schist 150 feet
thick, dipping 70° NW, contains prismatic sillimanite disseminated throughout the rock and coarse bladed
sillimanite in association with pegmatite veins. A channel sample, taken across the entire 159-foot ex-
posure, contained 8.0 percent sinks (Laboratory sample 687). The location would be ideal as a mine site
for combination open cut and quarry-face work into a steep hill to the southwest. Adequate dump area is
available in the steep drop to Jarrett Creek. Probably 150,000 cubic yards of sillimanite schist would be
available initially, with a possibility of extension in to the mountain to the southwest.
This same sillimanite zone reappears 1,000 feet to the northeast in a bank of the Wayah Gap road on
the north side of Jarrett Creek. The zone is less than 100 feet thick and possibly has a lower sillimanite
On the Wayah Gap-Aquone road, 1,000 feet east of its junction with the Rainbow Springs road, a
rather lean sillimanite schist is repeated several times by lateral folds. The zone apparently is an exten-
sion of the one previously described as being on the Jarrett Knob ridge. Additional exposures to the
northeast were not found.
26 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Except for a few small pieces of float 1,500 feet southeast of Jarrett Bald, sillimanite has not been
found between the Aquone-Wayah Gap road and Whiteoak Creek, a distance of nearly four miles. Along
the north fork of White Oak Creek, west of Licklog Gap, fibrous-massive sillimanite occurs as thick coatings
on pegmatite quartz stringers, and as veinlets in quartz-muscovite schist. Similar forms of sillimanite oc-
cur at intervals from the head of Gold Pit Branch to about 1,500 feet east of Burningtown Gap.
An apparent extension of the White Oak Creek occurrence crosses an abandoned road 4,000 feet east of
Burningtown Gap. Here fibrous and prismatic sillimanite is distributed in a 60-foot zone of a coarse-
grained series of alternating muscovite schist and quartz-biotite gneiss. Blades of coarse sillimanite have
been found in the creek bed but no concentrations were observed.
Prismatic sillimanite is found in a series of biotite schist zones, about 10 feet thick, on small tributary
branches of Younce Creek, three miles northeast of the head of Burningtown Creek, and large blades of
sillimanite occur in float of biotite gneiss about 2,000 feet east of Rattlesnake Knob at the head of Younce
Creek. It is doubtful that any of these occurrences have commercial value.
From Younce Creek the sillimanite belt widens to include a cross-strike distance of nearly two miles.
It is possible that some of the widening may be due to repetition by folding, but additional field data are
needed in order to evaluate fully the geologic structure. Sillimanite float and occasional outcrops of silli-
manite-bearing muscovite-biotite schist occur on Edwards Branch, Kelly Cove (Laboratory sample 740),
and Kelly Ridge (Plate 10).
Saldeer Deposits (Plate 10)
Saldeer Gap is situated 2.0 miles S 23° W from Etna Post Office. The main ridge cut by the gap trends
south toward Horton Knob and north to Raby Bend of the Little Tennessee River. A prominent spur
ridge trends westward to the Burningtown Creek road. Rock outcrops are scarce except in road cuts and
in places where more quartzose rocks occur. Narrow and lean sillimanite zones occur along the Burning-
town Creek road at 1,500 feet and 2,500 feet north of the mouth of Allen Branch, at the mouth of Kelly
Cove Branch, at 1,400 feet northeast of the mouth of Kelly Cove Branch, and at 2,500 feet southeast of the
mouth of Shope Cove Branch. These are occurrences of finely prismatic sillimanite disseminated in quartz-
sericite schist. Concentrations are not of sufficient volume and grade to be considered as sources of silli-
A sillimanite zone, nearly 1,000 feet wide, crops out on the Burningtown Creek road and on the point
of the spur ridge leading west from Saldeer Gap. The northwest side of the zone crosses the Burningtown
Creek road 900 feet southeast of the mouth of Kelly Cove Branch. The sillimanite, principally of the pris-
matic variety with minor quantities of fibrous and bladed varieties, is contained in a quartz-muscovite-bio-
tite-garnet schist. Outcrops of the rock are not apparent to the southwest of Burningtown Creek, nor to the
northeast of the spur ridge. A channel sample from a 40-foot outcrop near the southeast side of the zone
contained approximately eight percent sillimanite (Laboratory sample 709), and a representative chip sam-
ple from the northwest side of the zone was estimated to contain 10 percent sillimanite (Laboratory sample
710). From 15 to 20 percent of the sillimanite crystals have sericite inclusions. Zircon and garnet com-
posed about 10 percent of heavy mineral concentrates from the samples. Well over a million yards of silli-
manite-bearing rock would be available at the locality above road level in the event that the ore is found to
A smaller but richer sillimanite zone occurs on top of the ridge 1,200 feet north of Saldeer Gap, where
prismatic sillimanite is disseminated in coarse-grained, pegmatized quartz-muscovite-biotite schist, exposed
intermittently for a little more than 1,200 feet along a trend of N 60° E. The zone is 30 feet to 40 feet wide and
dips 60° NW. Two similar chip samples from across the entire width of the sillimanite zone contained an
average of 15.3 percent sillimanite, of which more than half had sericite inclusions (Laboratory sample 691).
Approximately 50,000 cubic yards of ore would be available initially. The location of the deposit on top of
a fairly sharp ridge and a spur would promote efficient mining by open cut methods. An unimproved road
crosses Saldeer Gap, 1,200 feet away, but new access would have to be made to the deposit. Water supply
in any large quantity is one-half mile from the deposit and 500 feet lower in elevation.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
28 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Sillimanite in garnet-quartz-muscovite schist is exposed irregularly as small outcrops and as float along
the ridge north from Saldeer Gap to Long Branch, thance along a spur ridge leading northeast toward Oak
Grove. Other exposures occur along a line from the point of Raby Bend southwest to the Burningtown
road and northeast to the Little Tennessee River at Etna, and at locations 700 feet west of the mouth of
Rose Creek and 2,000 feet SSW of the McCoy Bridge. None of these exposures are sufficiently rich for
Oak Grove Deposit (Plate 11)
Sillimanite occurs in a 300-foot zone on the north side of U. S. No. 64 at Oak Grove Cemetery, 2,000
feet northwest of Etna Post Office. The rock has bean weathered extensively, but appears to be a garnet-
mica schist. Small radiating bundles of prismatic sillimanite are well distributed throughout the exposure
and a few small blades of coarse sillimanite were found in a thin vertical pegmatite vein on the southeast
side of the zone. A chip sample collected from this outcrop contained an estimated 7.6 percent sinks (Lab-
oratory sample 698-C), all of which showed numerous sericite inclusions (See also sample 799). Indications
of an extension of the zone along the strike to the northeast or southwest are lacking, principally because
of concealing talus and alluvium.
Etna-Bradley Creek Deposits (Plate 11)
A probable extension of the Raby Bend sillimanite zone has been traced almost continuously northeast-
wardly from U. S. No. 64 at Etna Post Office to the southernmost gap of Mouse Mountain, a total distance
of 8,000 feet. The rock is quartz-muscovite-biotite-garnet schist, ranging in thickness from 50 feet to nearly
400 feet. The northeastern half of the exposed zone, from Bradley Creek to Mouse Mountain, appears to
be narrower and to contain less sillimanite than the southwestern half from Bradley Creek to the highway.
Sillimanite is present both as individual prisms and as radiating aggregates of prisms. The richer silliman-
ite is associated with pegmatization. Five channel samples, each representing 45 feet to 65 feet of rock,
were taken where the zone crosses the Bradley Creek road near the old Bradley place. Sillimanite content of
the samples ranged from 7 percent to 13.5 percent, and averaged 9.7 percent for the entire width (Labora-
tory samples 698-A, B, D, E, and 741). Associated heavy minerals include zircon, garnet, ilmenite, and
limonite; 50 percent of the sillimanite contained sericite inclusions. Rock outcrops along the secondary
road between Etna and Oak Grove appear to be as rich in sillimanite as those at Bradley Creek. A proba-
ble minimum of one-half million cubic yards of ore, averaging 10 percent sillimanite, is available between
Etna and Bradley Creek. Topographic conditions along the zone are favorable for open cut mining. Access,
water supply, electric power, and dumping facilities are excellent. A chip sample from the top of Mouse
Mountain contained 10.1 percent sillimanite (Laboratory sample 701), but the zone between Bradley Creek
and Mouse Mountain would have to be prospected by trenching or boring before volume and grade of the
sillimanite could be estimated.
Much of the Etna-Bradley Creek locality is overlain by alluvial material which conceals possible silli-
manite deposits. Indications of sillimanite have been found, however, in the following places : On U. S. No.
64 at 500 feet and 900 feet southeast of Etna Post Office, and on secondary roads at 2,400 feet N 70° E of
Etna, at 4,200 feet N 43° E of McCoy Bridge, and at 3,400 feet N 30° W and 2,500 feet N 5° E of the mouth
of Potts Branch. A chip sample from the latter location, representing 15 feet of rock contained 16.1 per-
cent sillimanite (Laboratory sample 690), indicating the desirability of additional prospecting.
Hickman Creek Deposits (Plate 11)
Rickman Creek is the western tributary of Matlock Creek, which joins with Cowee Creek one and one-
half miles northeast of Wests Mill Post Office. Three sillimanite horizons occur within a cross-strike dis-
tance of 700 feet along the Rickman Creek road, 5,000 feet north of Matlock Creek. None of the horizons
could be traced for more than a few hundred feet along the strike. The two northernmost zones are lean in
sillimanite but the lower zone is more encouraging. The latter is 65 feet in width and is composed of a quartz-
muscovite-biotite schist which strikes N 65° E, dipping nearly vertical. Sillimanite occurs in the schist as
disseminated prismatic and fibrous forms, also as the radiating bundles. Two channel samples were taken,
one representing 35 feet of rock (southeast side) contained 5.7 percent sillimanite, and one representing an
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
30 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
additional 30 feet section contained 2.6 percent sillimanite (Laboratory samples 688-A and B). In addi-
tion, from 5 percent to 10 percent combined kyanite, limonite, ilmenite, and zircon are present in the heavy
mineral faction. Probably one-half of the sillimanite has sericite inclusions. (Also see sample 702.)
Matlock Creek Deposits (Plate 11)
Sillimanite has been found on Matlock Creek above Rickman Creek at four localities. All exposures are
thin and low in sillimanite, and could not be traced beyond the discovery outcrops. One sample, taken
across 15 feet of rock at an abandoned house, 4,500 feet northeast of the mouth of Wests Branch, contained
9.7 percent sillimanite (Laboratory sample 700), all high in sericite inclusions.
Leatherman Deposits (Plate 12)
Indications of sillimanite have been found at widely scattered locations in the vicinity of Leatherman,
three miles east of Etna Post Office, but as yet only one zone appears worthy of prospecting. Immediately
below the mouth of Huckleberry Creek, a tributary of Beasley Creek, sillimanite occurs in quartz-muscovite
schist over a width of about 500 feet. A sample across 50 feet of a particularly rich horizon contained 19.5
percent sillimanite of low sericite content (Laboratory sample 711). Visual comparison indicated that an
additional 100 feet of section may contain more than 8 per cent sillimanite. Mining would not be feasible
to the northeast of the discovery exposure, but favorable conditions exist toward the southwest where eleva-
tions rise from 3,000 feet at the exposure to 3,300 feet on Cedar Cliff Ridge, a horizontal distance of 2,000
feet. Fresh rock is not exposed in the area, but firm rock probably would be encountered within 30 feet of
A zone of kyanite schist containing masses of pggmatite-type kyanite has been found from Mill Gap,
3,000 feet northwest of Leatherman, to the western slope of Panther Ridge, north of Beasley School. Pris-
matic and fibrous sillimanite is associated intimately with kyanite, but separation of a pure sillimanite con-
centrate would be difficult (See samples 689 and 746.)
Sillimanite-bearing rock occurs east of Huckleberry Creek to the top of Cowee Mountains, but quantity
and grade are not sufficient to justify sampling. Sillimanite schist float is scattered between Beasley and
Blazed Creeks and along the top of Panther Ridge. Sillimanite of the fibrous type also has been found in felds-
pathic garnet gneiss along the headwaters of Mica City Branch.
Cowee Church Deposits
Prismatic and fibrous sillimanite occurs in schist inclusion in granite gneiss on the western end of
Mason Mountain, just south of Cowee Church, about one mile south of Wests Mill Post Office. These de-
posits are just outside the southeastern limits of the main sillimanite belt. Two zones, 700 feet and 1500
feet south of Cowee Church, contain sillimanite in a zone about 150 feet thick. Concentrations of silliman-
ite, however, are restricted to 60-foot widths near the centers of the two zones (See samples 738 and 739).
Principal foliation of the rock strikes east-west and dips 85° SE. The more important associated minerals
are quartz, feldspar, sillimanite, sericite, and biotite. The rock is weathered but firm in exposures on
highway U. S. No. 64, but is concealed under several feet of overburden on the slopes of Mason Mountain
and in open land west of the highway. Prospect openings will be required in order to evaluate properly the
Greens Creek School Deposits (Plate 13)
A zone of about 150 feet of sillimanite-bearing migmatite extends for more than 1,000 feet northeast
of the Peewee Branch road, 2,500 feet northwest of Greens Creek School. A channel sample of 25 feet of
the best exposed part of the section contained 17.4 percent sillimanite (Laboratory sample 737). The rock
strikes N 55° E, dips 80° NW, and makes angle of about 45° with the Greens Creek road (See also sample
751). Ground elevation rises more than 300 feet in 1,400 feet along the zone to the northeast. A projec-
tion to the southwest lies generally in the bottomland of Greens Creek.
Sillimanite-bearing schist appears in a road cut immediately northeast of Greens Creek School at the
intersection of the Brushy Fork road. The zone is exposed at intervals for 2,000 feet northeast to its inter-
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
32 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
section with the electric transmission line, and a continuation farther northeast is possible, but an evalua-
tion is unsatisfactory without prospect openings.
Brook Branch-Sutton Branch Deposits (Plate 13)
The sillimanite zone exposed at Peewee Branch continues on and between Brook and Sutton Branches
at about 4,000 feet above their mouths. Sillimanite is associated with biotite in a quartz-muscovite schist
of moderate pegmatite content. Approximately 8 percent sillimanite is present in the rock over a width of
about 20 feet.
Sillimanite occurs on the Brook Branch road 2,200 feet above its mouth, but extension along the strike
was not found. At 800 feet and 400 feet above the mouth of Brook Branch, sillimanite zones of about 25-
foot widths occur in road cuts and have been traced intermittently for several thousand feet along the strike.
Exposures of sillimanite-bearing rock are not suffici3nt, however, to estimate economic value. A more
promising sillimanite zone, probably an extension of the occurrence on Brushy Fork, extends along Greens
Creek, across the mouth of Brook Branch, to 350 feet above the mouth of Sutton Branch. The sillimanite-
bearing portion of the feldspathic quartz-muscovite S2hist ranges in thickness from 10 feet to 60 feet. Local
dips range from 65° NW to vertical. The richer portion of the zone is about 1800 feet of strike length south-
west from Sutton Branch. A sample from the Sutton Branch end of the zone, representing 60 feet of sec-
tion, contained 20.0 percent sinks (Laboratory sample 752).
A rather lean zone of sillimanite-bearing schist, up to 200 feet wide, extends from 2,200 feet above the
mouth of Sutton Branch northeast to Cagle Branch. A 15-foot portion of the section was the only part of
the zone which seemed worthy of sampling, but prospect openings along the strike might reveal more prom-
Sillimanite occurrences are poor between Cagles Branch and the Tuckasegee River. Probable exten-
sions of the Peewee Branch-Sutton Branch zone and the Sutton Branch-Cagle Branch zone have been traced
to the northeast with difficulty. Almost continuous rock exposures across the sillimanite belt are present
in southwest road cuts of U. S. No. 64 along Tuckasegee River, but sillimanite of economic concentration is
Riverview Church Deposits (Plate 14)
A southwest-trending spur of Kings Mountain forms steep bluffs over the Tuckasegee River at from
1,000 feet to 2,000 feet northwest of Riverview Chur:h, 1.6 miles southwest of Sylva. Ledge rock is nearly
continuous across the 1,000 foot width of the spur, all containing sillimanite to some extent. At approxi-
mately equal intervals across the bluff, four zones of sillimanite concentrations, ranging from 20 feet to
60 feet in thickness, were sampled. The sillimanite-bearing rock can be studied at this locality both along
the strike and down the dip. Samples collected from this locality (Laboratory samples 717, 742-745) con-
tained between 6.4 and 19.6 percent sillimanite. Min'ng of the deposits would be aided by their occurrence
in steep bluffs, but operational troubles might result from the fact that dips in the rock vary from 65° NW
to 70° SE. ,
Intermittent exposures of sillimanite rock have been observed northeastward to the crest of Kings
Mountain (See sample 712) . Most of the ground is covered by talus, but it seems likely that firm rock would
be encountered at less than 20 feet beneath the surface. If sillimanite of economic concentration is found
to be continuous to the top of Kings Mountain, a comparatively unlimited supply of ore would be available.
Sylva Deposits (Plate 14)
Decomposed sillimanite-bearing schist is exposed in a cut at the end of a paved street 1,600 feet due
south of the Jackson County Courthouse in Sylva. Locally the foliation dips 70° SE and strikes N 35° E.
Intermittent outcrops indicate a zone width of about 100 feet but continuity along the strike could not be
verified. Small outcrops and some sillimanite schist float occur 2,000 feet east of this location, but none
seemed worthy of detailed prospecting.
A 30-foot zone of sillimanite-bearing feldspathic quartz-sericite schist, containing 11.3 percent silliman-
ite (Laboratory sample 736), crops out in a highway cut of U. S. No. 19-23, 450 feet north of its intersection
with N. C. No. 107. The zone crosses a small spur ridge and into a small stream valley at the northeast
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 35
corner of the Sylva city limits, a total outcrop distance of about 1,300 feet. Local strike and dip of the zone
is N 30° E, 85° SE. Rocks immediately northwest of the sillimanite have the appearance of decomposed
granite, probably representing a very strongly pegmatized mica schist. To the southeast of the zone the
rocks apparently are thoroughly weathered biotite-garnet gneiss. Sillimanite in the zone is of both fibrous
and prismatic form. Considerable alteration to sericite makes difficult any visual estimates of the silli-
manite content. Any mining at the locality would be comparatively easy because of the steep dip and mod-
erately high ground.
OTHER SILLIMANITE DEPOSITS
In addition to the two principal zones of sillimanite-bearing rocks in North Carolina, other sillimanite
occurrences are known and some of these occurrences are here described.
Grassy Ridge-Big Ridge Deposits
Although a well-defined sillimanite belt has not been traced northeast of Sylva, two additional sillimanite
localities have been found in a geologic setting different from that to the southwest. These localities are
near the Grassy Ridge mica mine, 2.1 miles S 36° E of Balsam Gap, in Jackson County, and at the Big Ridge
mica mine, 5.1 miles S 6° E of Waynesville, in Haywood County. The Big Ridge mine is four miles nearly
due east of the Grassy Ridge mine. The geology of the two localities is similar. Large mica pegmatites*
occur in quartz-mica schist and gneiss and quartz-garnet-biotite gneiss. Smaller masses of pegmatite are
present as pods, stringers, and laminae. Sillimanite is present in some of the quartz-garnet-biotite gneiss
as fibrous bundles and laminae.
Sillimanite was first noted on the Grassy Ridge mine access road 2,000 feet southwest of the mine (Lab-
oratory sample 656). Boulders of garnet-mica gneiss were found to contain appreciable sillimanite. A sam-
ple (No, 704) was analyzed and found to contain 20.6 percent heavy minerals, of which 70 percent was gar-
net, 25 percent was sillimanite, and 5 percent was combined rutile and ilmenite. Zircon, chromite, and
barite were present in trace quantities. During the field program, studies at the locality were hindered by
heavy foliage, so that exact data are limited. Apparently the sillimanite zone is rather thin and quite vari-
able in sillimanite content. Local strikes range from north-south to east-west, but the trend over a distance
of a mile or more seems to be about N 65° E. This was partially verified by the finding of small amounts
of sillimanite where the mine road crosses a small gap 2,800 feet southeast of Beetree Gap. Scattered silliman-
ite float occurs all along the strike between the gap and the Grassy Ridge mine, but no outcrops of high silli-
manite content were observed. The sillimanite in hand specimen is oriented along the major foliation of
the quartz-biotite-garnet gneiss, usually in augen-like masses up to one inch long and one-fourth inch thick.
Sillimanite appears to be present in large quantity, but microscopic examination shows considerable altera-
tion to, or replacement by, sericite and quartz.
Sillimanite has not been observed between the Grassy Ridge mica mine and the Big Ridge mica mine,
although chance of observation in that interval is poor because of a heavy foliage and limited outcrops. At
the Big Ridge locality the geology is similar to that at Grassy Ridge. Rocks including the mica pegmatite
are quartz-biotite-garnet gneiss which in places is quite schistose. Locally the rocks strike N 45° W to N 55°
W, and dip 40° to 70° SW. Sillimanite is most prominent in a cut of the mine access road between the main
adit opening and the principal surface openings of the mica mine. As at Grassy Ridge the sillimanite is
principally of the fibrous variety, arranged in small augen-like masses along the schistosity. Pseudomorphs
of quartz-sericite mixtures, after sillimanite, are quite noticeable even in hand specimens. Although sam-
ples have not been analyzed, it is doubtful that any of the rock thus far found has more than two or three
South Hominy Deposit
Sillimanite has been found at one locality in the South Hominy section of Buncombe County, 2.8 miles
north of Mount Pisgah. Strongly pegmatized quartz-garnet-biotite-muscovite gneiss and schist on the north-
ern end of Ripshin Mountain contains sillimanite in fibrous, prismatic, and coarsely bladed forms. Local
strikes range from N 65° E to N 80° W, and dips are from 65° SE to vertical. The sillimanitic zone appears
36 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
to be as much as 800 feet wide, as exposed on the crest of Ripshin Mountain and in test pits on property of
Julius West on the west slope of the mountain. Some kyanite is visible in overburden, and all forms of silli-
manite exhibit alteration to sericite and quartz. Exposures of the sillimanite rock are not sufficient to furnish
representative samples, so that sillimanite content of the material has not been determined. Additional
prospect openings will be required for a reasonable evaluation of the deposit. Field studies for several miles
east and west of the locality have not revealed additional sillimanite.
Asheville-Mount Mitchell Deposits
Two zones of sillimanite-bearing rock are known to occur near Asheville, Buncombe County, North
Carolina. Most of this area is in the Pisgah National Forest and is characterized by steep-sided mountains
and high peaks, Mount Mitchell being the highest peak east of the Mississippi with an elevation of 6,684.
The zone containing the more promising sillimanite deposits occurs approximately 2 miles north of
Asheville, striking roughly N 60° E. A small amount of geological work in the area indicates that the silli-
manite-bearing rock occurs in discontinuous bands from a point southwest of Asheville to or beyond the
Mount Mitchell area. Sillimanite was discovered on Mount Mitchell in the summer of 1949. Deposits of
schist containing sillimanite occur also on the northwest side of Jump Cove, 41/2 miles N 45° E of Ashe-
ville. At both of these localities the sillimanite occurs in a quartz-muscovite-biotite schist. Fine needle-like
crystals and sillimanite nodules up to 2 inches in diameter are present. A representative chip sample of
•schist taken from the northwest side of Jump Cove contained 17.3 percent sinks in heavy liquid (Labora-
tory sample 770).
Sillimanite-bearing rock crosses the Asheville Recreation Park 3 1 /-) miles southeast of Asheville, strik-
ing roughly parallel to the zone described as occurring north of Asheville. All of the outcrops examined
at this occurrence contain a low percentage of sillimanite and exhibit extensive alteration to sericite.
The field samples collected during this survey were processed in the North Carolina Minerals Research
Laboratory by Euan K. Greene, Student Assistant of the North Carolina Division of Mineral Resources,
and Earl C. Van Horn.
METHOD EMPLOYED FOR SAMPLE ANALYSES
Since the quantity of sillimanite in sillimanite-bearing rock cannot be determined accurately by visual
inspection, it was necessary to develop a quick and simple laboratory technique for analyzing the samples
for their sillimanite content.
The following procedure was used to determine the sillimanite content of samples collected by field
Sillimanite ore was crushed and screened at — 100 mesh. A charge of 5 grams of the sized ore was
placed in a 250 cc. beaker and enough concentrated hydrochloric acid added to cover the material. The mix-
ture was diluted with an equal volume of tap water and placed on a water bath at 100° C. At intervals of
5-10 minutes, the mixture was diluted with additional water to a total of about 150 cc. volume and for a total
leaching time of about 20 minutes.
The leached material was filtered and washed thoroughly with tap water and oven-dried for one hour
at 105° C. A four-gram charge of leached ore was placed in a 15 cc. centrifuge tube and the volume brought
to 12 cc. with Acetylene tetrabromide (specific gravity of 2.95). The mixture was shaken thoroughly to
obtain good dispersion of the mineral grains. The charge tubes were then placed in a Clay-Adams Safety
Head centrifuge, operated on an A.C. current and centrifuged three minutes at position 6 (about 3700
r.p.m.). Water was used in the metal tube holders to distribute pressure on the glass tubes.
The floats were removed by decantation and any middling grains which remained in suspension were
poured off with the floats. The floats and sinks were re-centrifuged separately with acetylene tetrabro-
mide to insure a thorough cleansing.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
The heavy liquor was filtered from the floats for re-use. The floats and sinks were separately washed
several times with previously used benzene and given two final washes with unused benzene. Washing was
best accomplished in a beaker rather than on filter paper. Following the benzene wash the two products
were washed several times with hot tap water. Filter bottles and fluted funnels were used for rapid filter
washing. After drying for one hour at 105° C, the material was weighed and percentages calculated on
the basis of the total weight before leaching. Sillimanite (specific gravity of 3.23) in its pure form sinks
readily in the heavy liquid but sericite (specific gravity 2.6-3.1), an alteration product of sillimanite some-
times is carried down with the sinks. Partial alteration to sericite may also cause small quantities of
sillimanite to appear in the floats. Any heavy minerals such as garnet, ilmenite, sphene, rutile, zircon,
etc. also dilute the sinks.
The use of a petrographic microscope was necessary in order to apply percentage corrections required
by dilution of the sinks by sericite and other heavy minerals and by carry-over of sillimanite to the floats.
After the various fractions had been separated and dried, the individual sample fractions were mixed thor-
oughly on a cutting cloth and a small sample transferred to a glass slide where the fraction was mixed in
a drop of immersion liquid, refractive index 1.65. Corrective factors were obtained where necessary by
means of a microscopic grain count, except that grain counts were not necessary where the sillimanite frac-
tion was quite pure.
In the event that sinks from acetylene tetrabromide contain too great a fraction of heavy minerals for
rapid microscopic counts, the sinks may be separated further by use of methylene iodide (specific gravity
3.25). The procedure for using the methylene iodide is identical with that used with acetylene tetrabromide
except that the sillimanite concentrates appear as floats.
AND LOCATION OF SAJ
Near Hollis, Rutherford County
Road cut sample from Cages Mountain, Caldwell County
Road cut at Fox's Orchard, Iredell County
Surface float sample, Cages Mountain, Caldwell County
Along Cedar Valley, Alexander County
Smith Cliff Deposit, Burke County
Fox's Orchard Deposit, eleven miles northeast of Taylorsville,
Cages Mountain Deposit, Caldwell County
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Sawmill Ridge, Caldwell County
Four miles south of Taylorsville, Alexander County
Highway No. 16 north side of Brush Mountain, Wilkes County.
Dudley Shoals I Deposit, about 2 miles west of Dudley Shoals
Dudley Shoals I Deposit, about 2 miles west of Dudley Shoals
Dudley Shoals III Deposit, about 2 miles west of Dudley Shoals
Dudley Shoals III Deposit, about 2 miles west of Dudley Shoals
Dudley Shoals II Deposit, about 2 miles west of Dudley Shoals
Dudley Shoals III Deposit, about 2 miles west of Dudley Shoals
Near Hollis, Rutherford County
Dudley Shoals I Deposit, about IV2 miles west of Dudley Shoals
Dudley Shoals II Deposit, about 1^2 miles west of Dudley Shoals
Dudley Shoals I Deposit, about 11/2 miles west of Dudley Shoals,
* Further discussion elsewhere in report.
DESCRIPTION OF THE HEAVY MINERALS IN SILLIMANITE ORE
Certain samples of the sinks were selected for microscopic study. The results of microscopic exam-
inations of the washed and dried sinks are listed as follows:
648-C from Hollis, Rutherford County, had 11.6 percent sinks in acetylene tetrabromide. Microscopic ex-
amination revealed the presence of sillimanite, sericite alteration, chromite, rutile, zircon and zircon in-
clusions. The sillimanite is estimated to be 96 percent of the sinks with slight sericite alterations.
SlLLIMANITE DEPOSITS OP NORTH CAROLINA 39
648-D road-cut samples from Cages Mountain, Caldwell County, had 14.6 percent sinks in acetylene tetra-
bromide. Microscopic examination revealed the presence of sillimanite, sericite alterations, biotite, py-
rite, zircon and chromite. There were very extensive sericite alterations of the sillimanite (about 60 per-
cent of the grain), which made up 90 percent of the sinks. Approximately five percent of the sinks were
completely altered to sericite. Biotite and pyrite made up 7-8 percent of the sinks.
648-E, road-cut sample, from Fox's Orchard, Iredell County, had 13.8 percent sinks in acetylene tetrabro-
mide. Microscopic examination revealed the presence of sillimanite, sericite alterations, garnet, epidote,
chromite and zircon inclusions. The sillimanite is estimated to be 30 percent of the sinks, the garnet 50
percent, chromite 10 percent, epidote 5 percent. The sericite alterations were very slight.
648-H, surface float sample, from Cages Mountain, Caldwell County, had 13.4 percent sinks. Microscopic
examination revealed the presence of sillimanite, sericite alterations, chromite, zircon inclusion, rutile
and hematite. Ninety-six percent of the sinks were sillimanite and sericite alterations of sillimanite. The
sericite alteration was very high with 3-5 percent of the sinks showing complete alteration.
648-M, surface float sample along Cedar Valley, Alexander County, had 12.0 percent sinks. Microscopic
examination revealed the presence of sillimanite, sericite alterations, chromite, zircon, zircon inclusions
and hematite. Ninety-four percent of the sinks were sillimanite or sericite alterations of the sillimanite.
The sericite alterations were extensive (about 60 percent of grain) with 3-4 percent completely altered
668-B, surface float sample from Cages Mountain, Caldwell County, had 14.8 percent sinks. Microscopic
examination revealed the presence of sillimanite, tourmaline, garnet, rutile, and ilmenite. Sillimanite,
about 50 percent of which is being altered to sericite, made up about 75-80 percent of the sinks. The ex-
traneous material was mostly tourmaline (6-10 percent).
668-C, surface float from Cages Mountain, Caldwell County, had 13.6 percent sinks. Microscopic examina-
tion revealed the presence of sillimanite, rutile and ilmenite. Sillimanite, about 50 percent being altered
by sericite, made up about 98-99 percent of the sinks. The extraneous material is rutile and ilmenite.
668-D, surface float sample from Cages Mountain, Caldwell County, had 12.2 percent sinks. Microscopic
examination revealed the presence of sillimanite, ilmenite, rutile, and zircon. Sillimanite, about 50-55
percent being altered to sericite, made up about 93-95 percent of the sinks.
668-F, surface float sample from Cages Mountain, Caldwell County, had 14.2 percent sinks. Microscopic
examination revealed the presence of sillimanite, ilmenite, a small amount of rutile and zircon, and a few
zircon inclusions. Sillimanite, about 50 percent being altered to sericite, made up about 95-97 percent
of the sinks.
669-3, surface float sample from Cages Mountain, Caldwell County, had 10.6 percent sinks. Microscopic
examination revealed the presence of sillimanite, pyrite, and biotite. Sillimanite, with high sericite al-
teration, made up about 90-93 percent of the sinks and about 4-5 percent of the sinks was pyrite.
669-5, surface float sample from Cages Mountain, Caldwell County, had 15.4 percent sinks. Microscopic
examination revealed the presence of sillimanite, ilmenite, pyrite, limonite, rutile, and zircon. Silliman-
ite, about 60 percent being altered to sericite, made up about 92-94 percent of the sinks, 2-3 percent was
ilmenite, and 2-3 percent was pyrite and limonite.
694, high grade sample from road cut 4 miles south of Taylorsville, Alexander County, had 18.8 percent
sinks. Microscopic examination revealed the presence of sillimanite, rutile, ilmenite, biotite, small
amount of garnet, zircon and zircon inclusions, and a few tourmaline inclusions. Sillimanite, about 5-7
percent being altered to sericite, made up about 94-95 percent of the sinks. Most of the extraneous mate-
rial was rutile (3 percent), and ilmenite.
695, chip road cut sample 4 miles south of Taylorsville on Highway 16, Alexander County, had 11.6 percent
sinks. Microscopic examination revealed that sillimanite, about 2 percent being altered to sericite, made
40 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
up about 95 percent of the sinks. Most of the extraneous material was ilmenite and rutile. There was also
some zircon and zircon inclusions and a few tourmaline inclusions.
696-A, road cut from Highway 16 on north side of Brush Mountain, Wilkes County, had 14.0 percent sinks.
Microscopic examination revealed the presence of sillimanite, ilmenite, rutile, zircon and zircon inclu-
sions, and a slight amount of magnetite, 90-92 percent of the sinks was sillimanite. Most of the extrane-
ous material was ilmenite (4-5 percent) and rutile (3 percent).
696-B, sample from hill above road cut of 696-A on the north side of Brush Mountain of Highway 16,
Wilkes County, had 13.4 percent sinks. Microscopic examination revealed the presence of sillimanite, il-
menite, rutile, garnet, zircon and zircon inclusions. Sillimanite, about 2 percent being altered to sericite,
made up about 90-92 percent of the sinks. The extraneous material was about 5 percent ilmenite, 1-2
percent rutile, 1-2 percent zircon.
707-A, Dudley Shoals Deposit I, Caldwell County, had 24.0 percent sinks. Microscopic examination reveal-
ed the presence of sillimanite, hematite, specular hematite, staurolite, zircon and zircon inclusions, garnet,
and tourmaline. Sillimanite, about 2 percent being altered to sericite, made up about 85-88 percent of the
sinks and most of it was needle or lath shaped. 7-9 percent was hematite and specular hematite, 1-2 per-
cent staurolite, 1 percent zircon and zircon inclusions.
707-B, Dudley Shoals Deposit I, Caldwell County, had 20.4 percent sinks. Microscopic examination reveal-
ed the presence of sillimanite, hematite, specular hematite, staurolite, small amount of garnet, zircon and
zircon inclusions. Sillimanite, about 1-2 percent being altered to sericite, made up about 90 percent of
the sinks. 6-8 percent of the extraneous material was hematite and specular hematite and about 2 per-
cent was staurolite. The sillimanite was mostly needle or lath shaped.
707-C, Dudley Shoals Deposit III, Caldwell County, had 19.8 percent sinks. Microscopic examination re-
vealed the presence of sillimanite, hematite, specular hematite, staurolite, garnet, zircon and zircon in-
clusions, 80-85 percent of the sinks was sillimanite with about 1 percent sericite alteration. Most of the
extraneous material was hematite and specular hematite.
707-D, Dudley Shoals Deposit III, Caldwell County, had 35.2 percent sinks. Microscopic examination re-
vealed the presence of sillimanite, hematite, specular hematite, staurolite, zircon and zircon inclusions,
and a small amount of tourmaline. Sillimanite, about 1-2 percent being altered to sericite, made up about
90 percent of the sinks. 7-8 percent of the sinks was hematite and specular hematite, and 2-3 percent was
708-A, road cut sample from Dudley Shoals Deposit II, Caldwell County, had 30.0 percent sinks. Micro-
scopic examination revealed the presence of sillimanite, hematite, specular hematite, staurolite, zircon and
zircon inclusions. Sillimanite, about 2 percent being altered by sericite, made up about 90-92 percent of
the sinks. The extraneous material was mostly hematite and specular hematite, and about 3 percent stau-
708-B, sample from hill to east of road cut (708-A), Caldwell County, had 20.8 percent sinks. Microscopic
examination revealed the presence of sillimanite, hematite, staurolite, small amount of garnet, zircon and
zircon inclusions. Sillimanite, about 5-7 percent being altered to sericite, made up about 85 percent of
the sinks, and was mostly needle-like. 9-10 percent was hematite, and 2-3 percent was staurolite.
713-2, float from ridge 5 miles N. E. of Hollis, Cleveland County, had 33.2 percent sinks. Microscopic ex-
amination revealed the presence of sillimanite, barite, rutile, ilmenite, hematite, zircon. Sillimanite
made up about 85-90 percent of the sinks. About 6 percent of the sinks was barite. Very little sericite
alteration was observed.
713-4, representatiive float from field 2 miles N. E. of Hollis, Cleveland County, had 12.8 percent sinks.
Microscopic examination revealed the presence of sillimanite, barite, rutile, ilmenite, hematite, magnetite,
and a few zircon inclusions. About 90 percent of the sinks was sillimanite with no alteration, 2-3 per-
cent of the sinks was barite.
SlLLIMANITE DEPOSITS OP NORTH CAROLINA 41
713-8, float from ridge 1 mile N. E. of Polkville, Cleveland County, had 24.8 percent sinks. Microscopic
examination revealed the presence of sillimanite, ilmenite, staurolite, barite, hematite, garnet and zircon.
About 90 percent of the sinks was sillimanite. There was very little sericite alteration.
713-10, float from field 3!/2 miles N. E. of Casar, Cleveland County, had 16 percent sinks. Microscopic ex-
amination revealed the presence of sillimanite, ilmenite, barite, staurolite, rutile, zircon, and a few zircon
inclusions. About 95 percent of the sinks was sillimanite with practically no alteration to sericite. Most
of the extraneous material was ilmenite.
713-11, sample from ridge and road cut 2.8 miles N. E. of Casar, Cleveland County, had 16.4 percent sinks.
Microscopic examination revealed the presence of sillimanite, kyanite, ilmenite, hematite, barite, garnet,
and staurolite. About 85 percent of the sinks was sillimanite. About 4-5 percent of the sinks was kyanite.
713-13, road cut sample and float from field 3!/2 miles north of Lawndale, Cleveland County, had 12 percent
sinks. Microscopic examination revealed the presence of sillimanite, kyanite, ilmenite, staurolite, barite,
and tourmaline. About 70 percent of the sinks was sillimanite, 20-25 percent of the sinks was kyanite.
ANALYSIS AND LOCATION OF SAMPLES FROM THE WARNE-SYLVA SILLIMANITE BELT
Downing Creek Road, Clay County
North side Downing Creek, Clay County
Cherry Mountain area, Clay County
Half mile northeast of Downing Creek Church, Clay County
Float sample from Grassy Ridge Mica Mine, Jackson County
Road cut at Oak Forest Church, Clay County
Junction of Pecky Wood Branch and Sap Sucker Branch, Clay County
Downing Creek area, Clay County
Rainbow Spring-Aquone Road, Macon County
Rickman Creek, Macon County
Near Leatherman, Macon County
Near west fork of Potts Branch, Macon County
North of Saldeer Gap, Macon County
Bradley Creek near Oak Grove Church, Macon County
Oak Grove area, Macon County
Matlock Creek north of Owl Knob, Macon County
Mouse Mountain, Macon County
Rickman Creek area, Macon County
Grassy Ridge Mica Mine, Jackson County
Burningtown Creek road, Macon County
Junction of Beasley and Huckleberry Creeks, Macon County
Southwest spur of Kings Mountain, Jackson County
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Laboratory Percent Percent
751* 13.0 12.7
♦Further discussion elsewhere in report.
Near Riverview Church on Tuckasegee River, Jackson County
North end of Sylva, Jackson County
Peewee Branch of Greens Creek, Jackson County
Near Cowee Church, Macon County
Southeast of Cowee Church, Macon County
Between Saldeer Gap and Kelly Cove, Macon County
Near Etna Post Office, Macon County
One mile southeast of Dillsboro, Jackson County
About one mile southeast of Dillsboro, Jackson County
One-third Mile northwest of Riverview Church, Jackson County
Shepherd Creek near Transmission line northwest of Leatherman,
Half mile northwest of Sutton Branch, Greens Creek, Jackson County
Near mouth of Sutton Branch of Greens Creek, Jackson County
DESCRIPTION OF THE HEAVY MINERALS IN SlLLIMANITE ORE
FROM WARNE-SYLVA BELT
Certain samples of the sinks were selected for microscopic study. The results of microscopic exam-
inations of the washed and dried sinks are listed as follows:
656, surface float sample near Grassy Ridge Mica Mine, Jackson County, had 32.9 percent sinks. Micro-
scopic examination revealed sillimanite, garnet, chromite, zircon and zircon inclusions and biotite. Ap-
proximately 80 percent of the sinks were sillimanite, 10 percent were garnet and 5 percent were chromite.
673, road cut sample near Oak Forest Church, Clay County, had 7.2 percent sinks. Microscopic examina-
tion revealed the presence of sillimanite sericite alterations, chromite, ilmenite, garnet, rutile, staurolite,
and zircon. The sinks were 75 percent sillimanite with high sericite alterations.
688-A, thirty foot channel sample from the east bank of the road on Rickman Branch, 1300 feet north of
transmission line, 2.0 miles east, northeast of the Etna Post Office, Macon County, had 16.2 percent sinks.
Microscopic examination revealed the presence of kyanite, sillimanite, ilmenite, small amount of specular
hematite and zircon. About 30-35 percent was sillimanite, 60-70 percent kyanite, and 4-5 percent il-
688-B, location same as 688-A, had 13.0 percent sinks. Microscopic examination revealed the presence of
kyanite, sillimanite, ilmenite, zircon, and a few zircon inclusions. About 75 percent was kyanite, 3-4 per-
cent ilmenite, and about 20 percent sillimanite with high sericite alteration.
689, grab sample from west bank of the road 1600 feet NNW of Leatherman, about 3.6 miles ENE of Etna
Post Office, Macon County, had 13.6 percent sinks. Microscopic examination revealed the presence of
kyanite, sillimanite, rutile, ilmenite, garnet, zircon, and zircon inclusions, about 90 percent was kyanite,
6-7 percent sillimanite. The extraneous material was mostly rutile and ilmenite.
690, grab sample from west road cut, 150 feet west of West Pork of Potts Branch, 5100 feet WNW of Pleas-
ant Hill Church, and 6600 feet ESE of Etna Post Office, Macon County, had 18.4 percent sinks. Micro-
scopic examination revealed the presence of sillimanite, garnet, ilmenite, zircon, and very few pieces of
kyanite and rutile. Sillimanite, about 20 percent being altered to sericite, made up 85-90 percent of the
sinks. Garnet and ilmenite made up most of the extraneous material.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 43
691-1, sample from ridge 1200 feet north of Saldeer Gap, 10,000 feet SSW of Etna Post Office, Macon County,
had 21.0 percent sinks. Microscopic examination revealed the presence of sillimanite, hematite, and iron
stained material, biotite, garnet, zircon and zircon inclusions, and very little kyanite. Sillimanite, about
70 percent being altered to sericite, made up about 90 percent of the sinks. Most of the extraneous mate-
rial was hematite and iron stained material. There was less than 1 percent kyanite.
691-2, (same location as 691-1) had 14.0 percent sinks. Microscopic examination revealed the presence
of sillimanite, hematite and iron stained material, garnet, zircon and few zircon inclusions, and very little
kyanite. Sillimanite, about 70-75 percent being altered to sericite — 1.2 percent complete alteration, made
up about 85-87 percent of the sinks. Most of the extraneous material was hematite and iron stained ma-
terial, 3 percent garnet.
698-A, road cut sample from Fred Bradley's place on Bradley Creek, 5,150 feet N.E. of Oak Grove Church,
and 5,500 feet NNE of McCoy Bridge, Macon County, had 10.8 percent sinks. Microscopic examination re-
vealed the presence of sillimanite, garnet, ilmenite, kyanite, zircon. Sillimanite, about 65-70 percent being
altered to sericite, made, up about 93 percent of the sinks. About one percent of the sinks was kyanite.
698-B, same locality as 698-A, had 11.8 percent sinks. Microscopic examination revealed the presence of
sillimanite, garnet, ilmenite, hematite, zircon and kyanite. Sillimanite, about 70 percent being altered
to sericite, made up about 92 percent of the sinks. About 1 percent of the sinks was kyanite.
698-D, same locality as 698-A, had 15.2 percent sinks. Microscopic examination revealed the presence of
sillimanite, garnet, ilmenite, hematite, pyrite, zircon and zircon inclusions, small amount of staurolite,
and kyanite. Sillimanite, about 60-65 percent being altered to sericite, made up about 88-90 percent of the
sinks. About 1-2 percent of the sinks was kyanite.
699, road cut sample from Oak Grove Community, between Oak Grove Cemetery and U. S. Highway 64,
Macon County, had 19.8 percent sinks. Microscopic examination revealed the presence of sillimanite, gar-
net, ilmenite, hematite, and kyanite. About 50 percent of the sinks was garnet and around 1-2 percent was
kyanite. Sillimanite, about 80 percent being altered to sericite, made up about 40-45 percent of the sinks.
700, cut from east bank of creek above spring, on Matlock Creek at second creek crossing of county road,
5000 feet north of Owl Knob, 3.2 miles ENE of Etna Post Office, Macon County, had 13.8 percent sinks.
Microscopic examination revealed the presence of sillimanite, kyanite, garnet, ilmenite, hematite, biotite,
and zircon. Sillimanite, about 40 percent being altered to sericite, made up about 70 percent of the sinks.
About 6 percent of the sinks was kyanite.
701, from Mouse Mountain 6,500 feet SSW of Grant Knob, and 9,100 feet WNW of Owl Knob, Macon County,
had 14.4 percent sinks. Microscopic examination re sealed the presence of sillimanite, hematite, garnet,
rutile, and kyanite. Most of the extraneous material was iron oxide. Sillimanite, about 50 percent being
altered to sericite, made up about 70 percent of the sinks. About 7 or 8 percent of the sinks was kyanite.
702, from Rickman Branch 4,800 feet N.W. of Owl Knob and 7,800 feet south of Grant Knob, Macon County,
had 32.8 percent sinks. Microscopic examination revealed the presence of sillimanite, garnet, ilmenite,
epidote and hematite. Sillimanite, about 30 percent being altered to sericite, made up about 96 percent of
the sinks. (Sample not representative; sample taken for geologic control.)
704, chip sample from road cut near Grassy Ridge mica mine, Jackson County, had 20.6 percent sinks.
65-70 percent of the sinks was garnet, 25-30 percent sillimanite, and about 5 percent was rutile and il-
710, grab sample from Burningtown creek road on point of ridge going due west from Saldeer Gap, Macon
County, had 11.2 percent sinks. Microscopic examination revealed the presence of sillimanite, hematite,
ilmenite, rutile, barite, and zircon. Sillimanite, about 15-20 percent being altered to sericite, made up about
85 percent of the sinks.
711, grab sample from confluence of Beasley and Huckleberry Creeks, Macon County, had 20.1 percent
sinks. Microscopic examination revealed the presence of sillimanite, kyanite, barite, garnet, ilmenite, and
44 SlLLIMANITE DEPOSITS OP NORTH CAROLINA
zircon. Sillimanite, about 5 percent being altered to sericite, made up about 97 percent of the sinks. Ky-
anite made up less than 1 percent of the sinks.
717, sample taken 200 feet N.E. of Tuckasegee River, 1800 feet NNE of Riverview Church, about 1.8 miles
S.W. of Sylva in Jackson County, had 13.6 percent sinks. Microscopic examination revealed the presence
of sillimanite, garnet, kyanite, ilmenite, staurolite, and zircon. About 47 percent of the sinks was silli-
736, channel sample from 30 ft. of section, 100 yards north of "Y" on U. S. 19 at intersection of first road
to right after passing N. C. 107 going north, in Jackson County, had 13.3 percent sinks. Microscopic ex-
amination revealed the presence of sillimanite, iron oxide, ilmenite, garnet, kyanite, zircon, tourmaline,
and staurolite. Very little alteration to sericite. About 85 percent sillimanite.
737, chip sample from 25 ft. of section 500 ft. north of mouth of Peewee Branch of Greens Creek, Jackson
County, had 18.3 percent sinks. Microscopic examination revealed the presence of sillimanite, iron oxide,
ilmenite, garnet, kyanite, and zircon. About 95 percent of the sinks was sillimanite and about 10 percent
of the sillimanite is being altered to sericite.
738, sample from 60 ft. of section on U. S. 64, 1 mile south of West Mill, 700 ft. SSE of Cowee Church, Ma-
con County, had 14.0 percent sinks. Microscopic examination revealed the presence of sillimanite, rutile,
garnet, and very little ilmenite, iron oxide, zircon ani zircon inclusions, biotite and kyanite. Most of the
extraneous material was rutile. About 98 percent of the sinks was clear sillimanite.
739, sillimanite schist representing 60 ft. section, on U. S. 64 one mile south of West Mill, 1500 ft. SSE of
Cowee Church, Macon County, had 15.1 percent sinks. Microscopic examination revealed the presence of
sillimanite, ilmenite, epidote, kyanite, rutile, garnet, zircon and a few inclusions; 85-90 percent of the
sinks was clear sillimanite.
740, sillimanite schist, 75 ft. of section, on point of ridge leading west from Saldeer Gap, 1500-1800 ft.
south of the mouth of Kelly Cove Branch, Macon County, had 16.3 percent sinks. Microscopic examina-
tion revealed the presence of sillimanite, rutile, iron oxide, and a little ilmenite. Most of the extraneous
material was rutile; about 97 percent (30 percent of which shows alteration to sericite) of the sinks was
741, sillimanite schist, 25 ft. of section, 200 ft. north of Etna Post Office, Macon County, had 6.7 percent
sinks. Microscopic examination revealed the presence of sillimanite, garnet, iron oxide, kyanite and a
small amount of ilmenite, zircon and rutile. About 96-97 percent of the sinks was sillimanite. Very high
alteration to sericite (40-50 percent).
742, sillimanite gneiss, 40 ft. of section, 6000 ft. SSE of Dillsboro, 1,200 ft. north of Riverview Church,
Jackson County, had 21.3 percent sinks. Microscopic examination revealed the presence of sillimanite, gar-
net, biotite, small amount of ilmenite and iron oxide. Most of the extraneous material was garnet. About
92 percent (5-10 percent alteration to sericite) of the sinks was sillimanite.
743, sillimanite gneiss 5,800 ft. SSE of Dillsboro, 1,500 ft. north of Riverview Church, Jackson County,
had 13.3 percent sinks. Microscopic examination revealed the presence of sillimanite, garnet, kyanite,
and a very small amount of staurolite, ilmenite, tourmaline, iron oxide, zircon, rutile and biotite. About
92-93 percent (about 10 percent being altered to sericite) of the sinks was sillimanite.
744, sillimanite gneiss 5,800 ft. SSE of Dillsboro, 1,500 ft. NNW of Riverview Church, Jackson County, had
14.0 percent sinks. Microscopic examination revealed the presence of sillimanite, garnet, epidote and a
small amount of ilmenite, zircon, biotite, and kyanite. Most of the extraneous material was garnet.
About 80 percent of the sinks was sillimanite.
745, sillimanite gneiss, 30 ft. of section, 5,600 ft. SSE of Dillsboro, 1,800 ft. NNW of Riverview Church,
Jackson County, had 18.0 percent sinks. Microscopic examination revealed the presence of sillimanite,
garnet and a small amount of ilmenite, staurolite, biotite and zircon. Most of the extraneous material was
garnet. About 90 percent of the sinks was sillimanite.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 45
746, kyanite-sillimanite gneiss from Shepherd Creek, 2.5 miles NE of West Mill, under transmission line
1,300 ft. NNW of Leatherman, Macon County, had 17.0 percent sinks. Microscopic examination revealed
the presence of kyanite, sillimanite, garnet, epidote, rutile and a very few zircon inclusions. About 5-7
percent of the sinks was sillimanite. About 90 percent of the sinks was kyanite. (Some alteration to
751, 2,200 ft. NW of mouth of Sutton Branch of Greens Creek, Jackson County, had 13.0 percent sinks. Mi-
croscopic examination revealed the presence of sillimanite, garnet, ilmenite, biotite, zircon. About 98
percent of the sinks was sillimanite. Very little alteration to sericite.
752, chip channel sample of sillimanite schist, 60 ft. of section, 350 ft. NW of mouth of Sutton Branch of
Greens Creek, Jackson County, had 20.0 percent sinks. Microscopic examination revealed the presence of
sillimanite, garnet, ilmenite, and a small amount of staurolite, zircon and iron oxide.
The following pages prepared by Mason K. Banks present a general summary of laboratory beneficia-
tion work as applied to ores from some of the deposits discussed in the geological section of this bulletin.
There has been no coordinated laboratory program in connection with the field survey so far as detailed
beneficiation test work is concerned. The results shown were obtained by sporadic test work, either on
ores of particular field interest, or whenever a request was made for small amounts of concentrates by
ceramic or refractory interests for research purposes, thus they do not represent results of a finished, pol- N
ished process, but are shown merely to indicate what is possible with some of the deposits located by the
By use of flotation (with oleic acid, sodium pyrophosphate, and sodium meta silicate) at fine sizes (at
least minus 100 mesh), all ores tested will yield high grade sillimanite concentrates, unless the sillimanite
was sericitized. Sericitization causes low recovery and low-alumina concentrates. In reducing the Fe 2 03
content of the concentrates to below 1.0 percent, it was often necessary to pass the concentrates through a
Franz ferro-filter. In samples where weathering had caused red iron oxide staining of the sillimanite par-
ticles, leaching with zinc hydrosulfite and sulfuric acid was effective in improving the color and also in
reducing the iron content of the concentrates.
Cages Mountain Deposit — Laboratory sample No. 648-H from Cages Mountain deposit, Caldwell Coun-
ty, contained 12.2 percent sillimanite by heavy liquid and microscopic assay, most of which was partially
altered to sericite. A small amount of graphite also was present, along with biotite, quartz, and trace
amounts of iron-titanium minerals. Since the minerals were locked down to fine sizes, flotation was the
method used with feed ground to minus 200 mesh.
Concentration of this ore was complicated by sericitization of the sillimanite which prevented the
successful preparation of a high-alumina concentrate, itbeing virtually impossible to selectively float un-
altered sillimanite particles from altered particles. Test No. 2 shows that the alumina content of the con-
centrates was improved by reducing the recovery drastically. This was accomplished by reducing the
amount of collector, thereby concentrating only the most readily floatable particles. The use of sericite
depressants was unsuccessful.
The preparation of low-iron concentrates was not particularly difficult. Control of pH was vital in
depressing iron minerals. A pH of 7.0 ± 0.3 in the rougher and first cleaner was necessary for effective
depression of iron minerals.
Results of two tests, both performed on feed ground to minus 200 mesh in a pebble mill are shown be-
low. Test 1 shows high recovery with low alumina due to sericite. Test 2 shows low recovery, with cor-
respondingly higher alumina.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Oleic Sodium Pyro-
Acid pH phosphate Emuosol
Pebble Mill. _. 1 .
Graphite Float 0.2
Sillimanite Rougher Float 1.9 6.9
Sillimanite Cleaner Float 0.2 7.0 0.5
Sillimanite Cleaner Float _.. 0.1 6.9 0.5
Sillimanite Cleaner Float 0.2 4.4
Results: ., ...... % Wt. AI2O3 Fe 2 03 SiCh
Sillimanite concentrates 14.2 56.4 0.75 39.0
Sillimanite Rougher Float.
Sillimanite Cleaner Float...
Sillimanite Cleaner Float.. .
Sillimanite Cleaner Float..
Concentrates similar to those shown in Test 1 are being studied by the Ceramic Engineering Depart-
ment, Clemson College, South Carolina.
Fox's Orchard Deposit — The laboratory samples from Fox's Orchard deposit, Caldwell County, were
designated Laboratory numbers 614, 615, 616 and 662. The ore contained an average of 12 percent of
sericitized sillimanite. In addition to the usual quartz and biotite, the ore contained graphite, pyrite, and
titanite. The more weathered portions contained much clay and some alum.
The same problem encountered in samples from the Cages Mountain deposit caused considerable dif-
ficulty in concentrating this ore. Sericitization of the sillimanite resulted in low alumina content of the
flotation concentrates. The presence of pyrite and alum caused rather high consumption of sodium meta-
silicate in obtaining the desired pH for flotation.
Emulsol XI was found to be a good collector for the graphite in the ore. Following graphite removal,
it was necessary to remove the pyrite with xanthate. After pyrite flotation, the sillimanite was floated
with an emulsion of oleic acid and Emulsol XI, using sodium pyrophosphate as an iron-mineral depres-
sant. The emulsion of oleic acid and Emulsol XI showed an improvement over straight oleic acid. It had
stronger collection properties and seemed to give improved iron-mineral rejection. Best results on Sam-
ple No. 615 (24 percent sillimanite) using the above-outlined procedure are as follows: Recovery — 85.5%;
A1 2 3 — 55.9%; Fe 2 3 — 2.0%; Si0 2 — 39.8%; Ti0 3 — 0.3%. The tests were performed on feed ground to
minus 200 mesh. The Fe 2 3 content of the concentrate could probably be lowered by passing the flotation
concentrate through a ferro-filter.
Smith Cliff Deposit — Several samples of ore from the Smith Cliff deposit near Morganton, Burke Coun-
ty, were tested. Sample No. 44 consisted of nodules picked up on the surface. Samples Nos. 46 and 307 were
nodular schist. Sample No. 47 was residual weathered schist. Sample No. 45 was the "needle" schist type.
Sample No. 538 was the sericitized schist type.
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 47
A minus 1/2 mcn P lus 20 mesh sample of crushed hand-picked nodules from Sample No. 44 which
analyzed A1 2 3 60.6%; Fe 2 3 0.7% Si0 2 37.0%, was submitted to the U. S. Bureau of Mines, Electrotech-
nical Laboratory, Norris, Tennessee, for refractory evaluation. Although this sample was not prepared
by commercially reproducible methods, it is probable that a similar product could be prepared by using
heavy media procedure, followed by magnetic separation to reduce the iron content. At the time this
work was done equipment was not at hand to attempt such a procedure.
Sample No. 307, after grinding to minus 200 mesh and floating with oleic acid, sodium metasilicate,
and sodium pyrophosphate, yielded the following products:
% Wt. % Sillimanite Distribution
Sillimanite Concentrates 18.9 94.0 73.5
Sillimanite middlings 6.3 45.0 11.5
Sillimanite trailings . - 50.5 5.0 10.3
Sillimanite slimes 24.3 5.0 4.7
Total 100 . 24 . 3 100 .
The above concentrate had the following analysis: A1 2 3 58.7%; Fe 2 3 3.0%; Si0 2 36.4%.
Sample No. 538 contained 11 percent partially sericitized sillimanite and graphite, quartz, muscovite,
biotite and garnet. Using the oleic acid, sodium silicate, sodium pyrophosphate float after an emulsol
graphite float on 35 mesh, 65 mesh, and 100 mesh feed, the following results were obtained:
Grind Recovery AI2O3 Fe2C<3 Si02 T1O2
35 mesh 61 48.1 6.8 38.0 1.0
65 62 • 57.9 3.3 36.2 0.7
100 59 58.3 3.4 35.6 0.7
From the above data, it was concluded that the ore should be ground to minus 100 mesh in order to
liberate the sillimanite. Accordingly, a minus 100 mesh flotation concentrate was made, passed through
a ferro-filter, then leached with zinc hydrosulflte and sulfuric acid. Results were as follows:
Recovery AI2O3 Fe203 S1O2 T1O2
The problem of high Fe 2 3 due to oxide stains and locked grains at fine sizes prevented preparation
of coarse concentrates from this sample. Minus 100 mesh concentrates, which would run 5-6 percent
weight of ore processed with approximately one percent Fe 2 3 , could be prepared by flotation, ferro-fllter-
ing, and leaching.
Dudley Shoals Deposits — A number of samples tested from the Dudley Shoals deposits are classified
as follows: (1) "schist type," containing 12-20 percent sillimanite disseminated in fine sizes in quartz-
biotite schist (Laboratory Nos. 694 and 730). (2) "soil type," containing 3-6 percent minus 16 mesh grains
of clean residual sillimanite and 3-12 percent plus 16 mesh iron-stained nodules of sillimanite (Labortory
Nos. 692-B to F, 731-A and B, 758 and 838). (3) "nodule type," consisting of nodules of sillimanite varying
in size from 14 inch up to four feet across which occur scattered through the soils of the Dudley Shoals
deposits (Laboratory Nos. 692-A and 757).
Type 1, "schist type," presents the problem, as do the other schist types, of flotation at minus 200
mesh. The sillimanite from these deposits, however, is not sericitized to any appreciable extent. Results
obtained from flotation of Laboratory No. 694, representing this type, were as follows: Recovery 76%;
A1 2 3 61.5%; Fe 2 3 0.7%; Si0 2 37.3%; Ti0 2 0.3%. The concentrate has both high alumina and low iron.
48 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
Type 2, "the soil type," offers possibility of production of both minus 16 mesh notation concentrates
and plus 16 mesh refractory material, although production of the refractory-sized material presents a very
difficult problem. Laboratory numbers 758 and 838 were used in these tests.
The minus 16 mesh was prepared as follows: The crude sample was screened on a 16 mesh vibrating
screen. The minus 16 mesh sands were scrubbed with caustic and deslimed at 325 mesh. The deslimed
sands were dried and screened on a 60 mesh Hummer screen. The minus 16 mesh plus 60 mesh portion was
discarded as it contained less than 1 percent sillimanite.
The minus 60 mesh deslimed sands, containing roughly 14 percent sillimanite, was floated with oleic
acid, sodium metasilicate, and sodium pyrophosphate. The rougher float was cleaned twice with oleic acid
and sulfuric acid. The cleaned froth product was then de-oiled and ferro-filtered three times. Metallurg-
ical results from this procedure are shown below:
Thirty pounds of the above concentrate were submitted to the Department of Engineering Research,
North Carolina State College, Raleigh, North Carolina for study.
The plus 16 mesh fraction of the "soil type" samples contained about 22 percent sillimanite in the
form of rather badly iron-stained nodules. A possible method of beneflciation would be to pass the mate-
rial through a heavy-media separating unit and then pass the sinks over a magnetic separator. The lab-
oratory did not have an adequate magnetic separator for such a procedure, but a simulation was perform-
ed as follows: A representative sample was placed in acetylene tetrabromide (specific gravity 2.96), and
the sillimanite removed as sinks and gangue minerals removed as floats. The sinks were then passed over
by a hand magnet to remove the most magnetic (high iron) portion. Results are shown below:
%Wt. AI2O3 Fe 2 03 Si02 Ti0 2
Floats.... 45 .
Magnetic sinks 25.6
Non-magnetic sinks.. 29.4 53.4 9.1 36.6 0.3
The hand magnet was not strong enough to remove all of the iron-bearing particles. Although the
simulation outlined above failed to produce low-iron refractory sillimanite, the procedure should not be
abandoned until it is tried using a suitable magnetic separator, which might lower the iron to the range
of refractory specifications.
Type 3, "the large nodule type," as represented by Samples No. 692-A, analyzed 59.3 percent A1 2 3 and
2.7 percent Pe 2 3 without beneflciation. No attempt has been made to reduce the iron at refractory size
range, but after grinding to minus 100 mesh, the material was passed through a ferro-filter and leached
with zinc hydrosulfite and sulfuric acid. The product from this treatment analyzed as follows: A1 2 3
—60.2%; Fe 2 3 — 1.1%; Si0 2 — 38.3%; Ti0 2 — 0.2%.
Large nodules in Sample No. 757 were crushed to minus % inch and the minus % inch plus *4 mcn
fraction contained: A1 2 3 — 58.2%; Fe 2 3 — 2.5%; Si0 2 — 38.3%; Ti0 2 — 0.6%. Hand-picked pieces from
this product analyzed A1 2 3 — 60.8%; Fe 2 3 — 0.49%; Si0 2 — 38.6%; Ti0 2 — 0.12%.
Grassy Ridge Deposit — Sample No. 704 was taken from a sillimanite schist outcropping on Grassy
Ridge, Balsam Gap, Jackson County. This is the only sample from the western sillimanite belt on which
any flotation tests were performed. The ore contained 5 percent sillimanite.
SlLLIMANITE DEPOSITS OP NORTH CAROLINA 49
Using feed ground to minus 200 mesh, with the conventional oleic acid-sodium metasilicate-sodium
pyrophosphate notation procedure, a recovery of 80 percent of the sillimanite was effected, having the
following chemical analysis: A1 2 3 62.6%; Fe 2 3 2.0%; Si0 2 34.8%; Ti0 2 0:3%.
This concentrate has a very high alumina content, although the iron is somewhat high. The alumina
content is an illustration of what can be done if the sillimanite is not sericitized. All of the sillimanite
in this sample was clean and fresh with no sericite alteration.
Given a schist containing 15-20 percent sillimanite which is not appreciably sericitized, a minus 100
mesh concentrate containing 60 percent or more A1 2 3 and 1.0 percent or less Fe 2 3 may be produced by
froth notation. It is sometimes necessary to follow notation with leaching and ferro-filtering. Whether
or not there is a market for a high grade sillimanite concentrate in the minus 100 mesh size range is not
There is a market for material in the minus % inch plus 8 mesh size range for refractory use. Un-
fortunately, only the Dudley Shoals and Smith Cliff deposits contain sillimanite in large enough crystal
sizes to make refractory grog, and both of these contain a high percentage of iron, which occurs as oxide
stain and as locked iron-bearing minerals. The only successful method of reducing the iron was achieved
by first grinding the nodules to minus 100 mesh, the reduction in size thus destroying their usefulness as
refractory material. A method which might be practicable for concentration of refractory material from
nodular sillimanite would be the use of heavy-media, followed by magnetic separation of the sinks to re-
duce the iron content. This method was simulated and results shown under the Dudley Shoals section.
The iron content was still far too high, but lack of a suitable magnetic separator was partly responsible.
Summarizing, a minus 100 mesh flotation concentrate of excellent grade can be produced from almost
any unsericitized sillimanite schist. However, production of a refractory material still remains an unsolv-
ed problem due both to lack of suitable prospected deposits and, consequently, lack of opportunity to de-
velop a suitable beneficiation method.
PROSPECTING, MINING, AND RESERVES
In the search for sillimanite certain factors should be considered. In reconnaissance surveys unusual
characteristics of soils are important. As a rule, soil underlain by sillimanite-bearing rocks contain frag-
ments of only partially weathered rock along with the soil and is characterized by a "pebbly" appearance.
In the Cliffside-Elkin belt the sheared character of the sillimanite-bearing rock imparts a "fragmental"
Anyone interested in prospecting for sillimanite should not rely too greatly upon the abundance of
sillimanite float as an indication of a large underlying deposit. 14 The sillimanite ore has a much greater
resistance to physical and chemical weathering than does the country rock, thus frequently a series of
narrow zones with thick intervening barren zones may, upon weathering, leave numerous fragments of
sillimanite in the soil. This leaves the erroneous impression that the entire area is underlain by silli-
Once a sillimanite occurrence has been located, perhaps the best method of preliminary prospecting
is by pits and trenches which would serve to outline the deposit. These openings should, where possible,
be of a cross-cutting nature and extend to sufficient depth to expose rock in place, but not necessarily
fresh, unweathered rock. Channel samples should be obtained from the occurrence and the sillimanite
content and characteristics of the sillimanite concentrate studied. If, at this stage, the deposit still shows
promise, core drilling should be undertaken.
Before the start of active mining, large samples of ore should be processed through the pilot plant
stage in order to determine their milling characteristics, since various sillimanite ores react differently
to any one beneficiation process.
50 SlLLIMANITE DEPOSITS OF NORTH CAROLINA
The North Carolina sillimanite deposits generally occur in low, elongated ridges, thus open cut quar-
rying or mining can be used. In the Cliff side-Elkin belt, although there is considerable variation in the
amount of overburden, firm rock generally will be encountered at depths between 15 and 30 feet. At some
of the deposits the overburden material contains sufficient sillimanite to justify its processing to recover
the mineral. Depth to fresh unweathered rock in this belt is not known. In the Warne-Sylva belt there
is no apparent topographic expression related to the sillimanite rock, however, since the sillimanite occurs
in mountainous terrain, open cut mining methods can be used. Overburden thickness ranges from a foot or
two up to as much as 75 feet, depending upon the distance from streams.
Because of the variable concentration of the sillimanite ore, more or less selective mining should be
practiced in any mining operation. Granite and pegmatite intrusions in the deposits which exceed a few
feet in thickness can be removed from the ore before milling, however, because of the lit par lit nature of
the pegmatite intrusions and the sillimanite-bearing rock a considerable amount of barren rock will have
to be milled.
Although no accurate estimates as to the reserves are available in either of the two belts, the depth to
which sillimanite extends being unknown; certain general statements about reserves can be made. In the
Cliffside-Elkin belt, listed in order of size, the principal deposits are: Smith-Cliff deposit, Cages Mountain
deposit, and the Dudley Shoals deposits. This estimate of reserve is based on the size of the deposit and not
on the sillimanite content.
In the Warne-Sylva belt, as has been pointed out previously, zones of sillimanite-bearing rock up to sev-
eral hundred feet thick occur, but, due to discontinuous exposures, lengths of the various occurrences are un-
known. It is believed, however, that those deposits described in detail in the chapter on "Descriptions of
Deposits" contain sufficient reserves to permit the operation of a plant of moderate size.
POSSIBILITIES FOR DEVELOPMENT
The annual consumption of the high alumina silicate minerals which include kyanite, andalusite, and
dumortierite, can be estimated only roughly, as many of the producers of these minerals are also consumers,
thus true production figures are not available. It is believed that the preparation of a high quality concen-
trate would permit marketing of more than 20,000 tons of sillimanite per annum. Domestic high alumi-
nous materials suited for high temperature use are distributed quite widely, except for some of the more
promising materials such as massive topaz from Chesterfield, South Carolina, 1 and massive kyanite similar
to Indian kyanite from Georgia 2 which occur in limit ad quantity. The use of andalusite and dumortierite
from the western states has been restricted to specialized production, since the location and character of the
deposits are such that large production is impractical. Crystalline bladed kyanite, disseminated throughout
metamorphic schists and gneisses, is abundant in the southeastern states. Its use in refractories requiring
coarse grog has been limited however, because of its decrepitation characteristics during calcination.
No domestic deposits of sillimanite have been worked on a commercial scale. The known deposits of
domestic sillimanite, like the kyanite in the southeastern states, occurs as disseminated crystals in meta-
morphosed rock, thus before use it will require benefication. In comparison with kyanite, sillimanite does
not require pre-calcination before use, and the grain s'.ze of the sillimanite particles is limited to that size
which can be obtained during extraction of the mineral.
Similar conditions are involved in the mining of sillimanite and kyanite, and mining costs are compar-
able. The only two active producers of kyanite in the southeast at present, located in Virginia and South
Carolina, recover kyanite which occurs disseminated in quartzite. Kyanite content of those ores now in pro-
duction range between 20 and 35 percent, which is probably higher than any of the deposits discussed in
this report. Local sillimanite contents, however, range up to 35 percent.
The only successful method developed to date to recover the sillimanite content of sillimanite-bearing
rock which occurs in North Carolina has been flotation. Alteration of the sillimanite to sericite has caused
considerable variation in the reaction of sillimanite ores to flotation. Also the sericite, which is extremely
difficult to separate from the sillimanite, lowers the ref ractiveness and is evident in the chemical composition
SlLLIMANITE DEPOSITS OF NORTH CAROLINA 51
of the concentrate as the aluminum content is lowered with a corresponding increase in the potassium con-
In order to exploit the sillimanite deposits in North Carolina, it will be necessary to produce a concen-
trate at approximately the. same or slightly higher cost as compared to kyanite and, if the cost is higher,
with corresponding improvements in thermal characteristics. It is thought that the ore should contain in
excess of 10 percent sillimanite with a minimum of saricite alteration. A deposit, in order to be commercial,
should have a tonnage reserve of at least 1,000,000 tons, and should be situated near rail transportation and
a dependable water suppply.
Because of the disseminated nature of the sillimanite ore, a large amount of sillimanite in fine sizes will
result from any operation. It will be necessary to develop a market, probably in the ceramic porcelain in-
dustry, to utilize this finer sillimanite fraction. A certain amount of iron oxide is present in all of the con-
centrates prepared to date. In order to produce a high quality concentrate, this iron contamination should
be kept to a minimum. In order to obtain this condition, it may be necessary to resort to magnetic separa-
tion and/or acid leaching of the concentrate.
1. Fries, C, Jr., "Topaz Deposits Near the Brewer Mine, Chesterfield County, South Carolina," U. S. Geol.
Survey Bull. 936-C, 1942.
2. Furcron, A. S. and Teague, Kefton H., "Sillimanite and Massive Kyanite in Georgia (A Preliminary
Report)," Georgia Geological Survey, Bulletin 51, 1945.
3. Hunter, Charles E. and White, William A., "Occurrences of Sillimanite in North Carolina," Information
Circular 4, North Carolina Department of Conservation and Development, February 1946.
4. Keith, Arthur, U. S. Geological Survey, Geologic Atlas, Washington Folio, No. 70, 1901.
5. , U. S. Geological Survey, Geologic Atlas, Nantahala Folio, No. 143, 1907.
6. Moore, George E., Jr., "Structure and Metamorphism of the Keene-Battlesboro, New Hampshire Area,"
Bulletin Geological Society of America, Volume 60, No. 10, October 1949.
7. Pratt, J. H. and Lewis, J. V., "Corundum and the Peridotites of Western North Carolina," North Caro-
lina Geological Survey, Volume 1, 1905.
8. Pratt, J. H., "Zircon, Monazite, and Other Minerals Used in the Production of Chemical Compounds
Employed in the Manufacture of Lighting Apparatus," North Carolina Geologic and Economic Survey,
Bulletin 25, 1916.
9. Prindle, Louis M. and Others, "Kyanite and Vermiculite Deposits of Georgia," Georgia Geological Sur-
vey, Bulletin 46, 1935.
10. Ramberg, Hans, "The Faces Classification of Rocks," Journal of Geology, Volume 57, No. 1, January 1949.
11. Robinson, Gilbert C, "The Use of South Carolina Sillimanite in a Porcelain Composition," presented at
the Annual American Ceramic Society Meeting in New York, April 1950.
12. , "The Resistance of a Sillimanite Porcelain to Abrupt Heating and Cooling Schedules,"
presented at the Annual American Ceramic Society Meeting in New York, April 1950.
13. Smith, L. L., "Sillimanite in South Carolina," Economic Geology, Volume XL, No. 4, June-July 1945.
14. Teague, Kefton H., "Sillimanite in the Southeast," Trans. AIME, Vol. 187, July 1950.
15. Turner, Francis J., "Mineralogical and Structural Evolution of the Metamorphic Rocks," Geological So-
ciety of America, Memoir 30, 1948.
16 Weiss, Judith, "Wissahickon Schist at Philadelphia, Pennsylvania," Bulletin Geological Society of
America, Vol. 60, No. 10, October 1949.